System and method for automatic in-line plunger depth measurement

Abstract

The automated inspection system includes a sensor system configured to generate a plurality of syringe scans by scanning each of a plurality of syringes, each of the plurality of syringe scans indicating a distance relative to the sensor. The automated inspection system also includes one or more processors configured to analyze, for each of the plurality of syringes, a respective syringe scan to determine (i) a first distance to a first portion (e.g., a flange) of the syringe and (ii) a second distance to a second portion (e.g., a plunger) of the syringe, and to calculate a distance between the first portion and the second portion based on the first distance and the second distance.

Description

[Technical field]

[0001] This application relates generally to quality control techniques, and more specifically to techniques for measuring distances between parts of a syringe (eg, for measuring the depth of a syringe plunger). [Background technology]

[0002] The manufacture of various pharmaceutical drugs and the containers that house those drugs requires strict quality control. For liquid-based drugs in syringes, one such quality control measure includes inspecting each syringe to ensure that the plunger (e.g., rubber piston or stopper) is at the proper depth within the syringe barrel. Plunger depth is typically measured as the distance between the top of the syringe flange and the top of the plunger while the syringe is in an upright position with the needle pointing down. Plunger depth is typically checked during the filling process and, for combination / autoinjection devices, a second check is made prior to assembly.

[0003] At the time of filling, in-process control (IPC) is employed to ensure that the settings of the automated filling line are correct and that the filling machine operates according to the settings. For automated injection devices (e.g., SureClick® or other combination devices), inspections are repeated after filling and during transportation, especially before assembly, as the plunger may move / subside if the ambient air pressure changes. In automated injection devices, the plunger depth may be out of the specified depth range, causing the automated injector to malfunction or deliver an incorrect dose. For example, if the plunger is positioned too low in the syringe (i.e., toward the needle), the spring-activated piston used by the automated injection pen may travel a longer distance before contact, thereby collecting excess kinetic energy. This may generate a shock wave into the glass body of the syringe upon contact with the plunger, resulting in glass breakage near the tapered end of the needle. The plunger position may also affect other quality-related aspects, such as container closure integrity and sterility. It is therefore important that robust processes are used to ensure that plunger position falls within predetermined specifications for both the filling and assembly stages.

[0004] Traditionally, plunger depth measurements have been performed in a manual process (e.g., using calipers or optical comparators). More recently, automated visual inspection (AVI) solutions have been developed that use machine vision to mitigate the inconsistencies and inefficiencies associated with manual processes. However, these AVI techniques can face their own drawbacks. For example, plunger depth measurements typically require a non-destructive and non-contact approach, so AVI techniques developed to date measure plunger depth using a camera that looks through the cylindrical sidewall (i.e., perpendicular to the central / long axis of the syringe). This presents a problem when the syringe is placed in certain types of containers (e.g., syringe "tubs"), as is common during and after the filling stage, because the container prevents imaging from a side perspective. Thus, these AVI techniques generally require the syringe to undergo additional handling (e.g., being removed from a standard syringe tub) beyond that already required for the filling and assembly lines. This additional handling can come with other drawbacks, such as reduced throughput, more expensive testing equipment, increased measurement variability due to tolerances in how each syringe is handled / held, increased opportunity for contamination, and additional mechanisms that can become failure points.

[0005] AVI technology can also be problematic in applications that do not utilize syringe tabs, such as when syringes are inspected while in a round tray, star wheel, or linear assembly line, particularly where the system geometry may not be suitable for integration of a side view camera and / or the machine vision technology introduces image processing delays that can make the plunger depth inspection point a bottleneck in the manufacturing process. Summary of the Invention [Means for solving the problem]

[0006] To address some of the aforementioned shortcomings of current manual and AVI implementations, the embodiments described herein employ a non-destructive, non-contact measurement technique that scans a syringe using a proximal end viewpoint (i.e., facing the end of the syringe opposite the syringe needle) to inspect the plunger depth level of the syringe with high accuracy and high throughput. In some embodiments, for example, the scan is generated by scanning the syringe using one or more sensors that pass over the proximal end of the syringe upright (i.e., needle facing down) or redirect the light path to the proximal end of the syringe using a mirror. The sensor can be, for example, a sensor with a passive sensor head, such as a confocal chromatic sensor (where different wavelengths are focused at different distances from the sensor head) or a "single focal depth" sensor (which has a single focal length and the sensor head is moved towards or away from the sample to find or approximate the distance that results in maximum focus). The scan can be a one-dimensional depth profile or a two-dimensional image / array of depth measurements depending on the sensor's scan pattern (e.g., linear or raster scan). The sensors can generate measurements at very high speeds (e.g., 25,000-70,000 times per second) and the associated processing can be relatively fast / low complexity (e.g., compared to machine vision image processing), thereby preventing substantial delays or bottlenecks. Speed ​​can be further increased by arranging multiple sensors in a configuration that allows the sensors to scan different syringes in parallel, or that allows different sensors to scan different parts of the same syringe. Furthermore, the proximal end viewpoint of the sensor allows plunger depth to be inspected in some embodiments without deviating from normal handling / transport of the syringe (e.g., without removing the syringe from the syringe tub that holds it upright). This allows for greater flexibility in the design of the inspection system. For example, a standalone inspection system employing these techniques (such as a tub-based system) can be placed virtually anywhere in the production chain.More generally, although the techniques described herein are described primarily with respect to detecting plunger depth (e.g., the distance between a flange and a plunger), these techniques may be applied to determine the distance between any two parts of a syringe, so long as the sensor is capable of individually detecting / determining the distance between the sensor (e.g., the sensor head) and each of these two parts.

[0007] Those skilled in the art will appreciate that the figures described herein are included for illustrative purposes and are not intended to limit the disclosure. The figures are not necessarily to scale, with emphasis instead being placed on illustrating the principles of the disclosure. It should be understood that in some instances, various aspects of the described implementations may be shown exaggerated or enlarged to facilitate understanding of the described implementations. In the drawings, like reference numbers generally refer to functionally similar and / or structurally similar components throughout the various views. [Brief description of the drawings]

[0008] [Figure 1] FIG. 1 is a simplified block diagram of an exemplary automated inspection station capable of implementing the syringe depth inspection techniques described herein. [Diagram 2] FIG. 1 is a perspective view of an exemplary automated inspection station for inspecting syringes in tubs. [Diagram 3] 1 illustrates one way in which plunger depth can be measured. [Figure 4] 1 illustrates exemplary orientations in which syringes may be held, including one syringe held with an angular displacement. [Figure 5A] 2 illustrates an exemplary confocal chromatic sensor that may be used as a sensor in the automated inspection station of FIG. 1. [Figure 5B] 2 illustrates an exemplary single focus depth sensor that may be used as a sensor in the automated inspection station of FIG. 1. [Figure 5C] 2 illustrates an exemplary single focus depth sensor that may be used as a sensor in the automated inspection station of FIG. 1. [Figure 6A] 2 shows an example depth profile and depth image, respectively, that may be generated using the sensor system of FIG. 1; [Figure 6B] 2 shows an example depth profile and depth image, respectively, that may be generated using the sensor system of FIG. 1; [Figure 7] FIG. 2 is a flow diagram of an exemplary algorithm that may be implemented by the automated inspection station of FIG. 1. [Figure 8] FIG. 1 is a perspective view of an exemplary automated inspection station for inspecting syringes in tubs using multiple sensors. [Figure 9A] 1A-1D are various perspective and overhead views of an exemplary automated inspection station for inspecting syringes on a linear transport mechanism. [Figure 9B] 1A-1D are various perspective and overhead views of an exemplary automated inspection station for inspecting syringes on a linear transport mechanism. [Figure 9C] 1A-1D are various perspective and overhead views of an exemplary automated inspection station for inspecting syringes on a linear transport mechanism. [Figure 9D] 1A-1D are various perspective and overhead views of an exemplary automated inspection station for inspecting syringes on a linear transport mechanism. [Figure 10A] 1A-1D are various perspective and overhead views of an exemplary automated inspection station for inspecting syringes on a linear transport mechanism using multiple sensors. [Figure 10B] 1A-1D are various perspective and overhead views of an exemplary automated inspection station for inspecting syringes on a linear transport mechanism using multiple sensors. [Figure 10C] 1A-1D are various perspective and overhead views of an exemplary automated inspection station for inspecting syringes on a linear transport mechanism using multiple sensors. [Figure 10D] 1A-1D are various perspective and overhead views of an exemplary automated inspection station for inspecting syringes on a linear transport mechanism using multiple sensors. [Figure 11A]1A-1D are various perspective and overhead views of an alternative automated inspection station for inspecting syringes on a linear transport mechanism using multiple sensors. [Figure 11B] 1A-1D are various perspective and overhead views of an alternative automated inspection station for inspecting syringes on a linear transport mechanism using multiple sensors. [Figure 11C] 1A-1D are various perspective and overhead views of an alternative automated inspection station for inspecting syringes on a linear transport mechanism using multiple sensors. [Figure 11D] 1A-1D are various perspective and overhead views of an alternative automated inspection station for inspecting syringes on a linear transport mechanism using multiple sensors. [Figure 12A] 1 illustrates an exemplary automated inspection station in which multiple sensors are positioned along a star wheel at different radial offsets. [Figure 12B] 1 illustrates an exemplary automated inspection station in which multiple sensors are positioned along a star wheel at different radial offsets. [Figure 12C] 1 illustrates an exemplary automated inspection station in which multiple sensors are positioned along a star wheel at different radial offsets. [Figure 13] 13 is a table showing calculated plunger depths for ten sample syringes across a range of different flange orientations determined using a confocal chromatic sensor. [Figure 14] 1 is a plot comparing plunger depth determined using a confocal chromatic sensor with plunger depth determined using an optical comparator. [Figure 15] FIG. 1 is a flow diagram of an exemplary method for automated inspection of a syringe. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] The various concepts described introductory above and discussed in more detail below can be implemented in any of numerous ways, and the described concepts are not limited to any particular implementation manner, examples of which are provided for illustrative purposes.

[0010] FIG. 1 is a simplified block diagram of an exemplary automated inspection station 100 that can inspect syringes 102 for proper plunger depth. More generally, in various embodiments, the automated inspection station 100 can be used to inspect syringes 102 to determine the distance between two portions of each syringe 102. The automated inspection station 100 can be a separate station or a single station in a line equipment having, for example, multiple automated inspection stations (e.g., one or more camera-based AVI stations for detecting other types of defects in addition to the automated inspection station 100). Each of the syringes 102 includes at least a barrel (e.g., a cylindrical tube), a plunger (which may also be referred to as a stopper or piston) disposed within the barrel and arranged to move / slide within the barrel, and a flange at the proximal end of the syringe. The syringe also includes a needle or other discharge orifice at the distal end of the syringe. Exemplary syringes are shown in FIGS. 3-5. The barrel and flange may be constructed of glass (e.g., borosilicate glass), plastic, and / or any other suitable material. The plunger may be constructed of rubber, plastic, and / or other suitable material. The needle (or other discharge orifice) may be constructed of metal, plastic, and / or other suitable material. Any of the components of the syringe 102 may be constructed using transparent, translucent, and / or opaque materials.

[0011] In some embodiments, each of the syringes 102 is held in an upright position (e.g., in a syringe tub or rondo tray or in an individual syringe holder on a star wheel or linear filling or assembly line, etc.). As the term is used herein, an "upright" position of the syringe is one in which the proximal / flange end of the syringe faces up and the distal / orifice end of the syringe faces down. The syringes 102 may be provided to the automated inspection station 100 by a previous inspection station (e.g., in the same line equipment as the automated inspection station 100) or manually, etc. The automated inspection station 100 includes a syringe transport mechanism 110 that receives the syringes 102 and moves the syringes 102 to a position suitable for scanning. For example, if the syringes 102 are held in a tub or rondo tray, the syringe transport mechanism 110 may include a conveyor belt on which the tubs are stacked and a drive mechanism (e.g., a motor) that moves the tubs to the conveyor belt. As another example, if the syringes 102 are not held in a common container, the syringe transport mechanism 110 can include a star wheel and a drive motor or pneumatic mechanism that rotates the star wheel in precise increments. As yet another example, the syringe transport mechanism 110 can be a linear transport mechanism. Various examples of automated inspection stations that can be used as station 100 are shown in Figures 2 and 8-12, which are described in further detail below. In some embodiments (e.g., if the tab of the syringe 102 is manually placed into the automated inspection station 100), the automated inspection station 100 does not include a syringe transport mechanism 110.

[0012] The automated inspection station 100 also includes a sensor system 112 having at least one sensor that scans the syringes 102 from a proximal viewpoint once the syringe transport mechanism 110 has properly positioned the syringes 102 (or once the syringes 102 have been manually positioned). That is, a sensor of the sensor system 112 scans the proximal end of each of the syringes 102. For example, if the syringes 102 are in an upright (e.g., needle down) position, the sensor may be positioned directly above a given syringe when scanning that syringe. The sensor may achieve a proximal viewpoint directly through the positioning of the sensor (i.e., such that the optical path between the sensor and the syringe 102 is a straight line) or using an intervening optical system (e.g., one or more mirrors redirect the optical path between the sensor and the syringe 102, such as the arrangements shown in FIGS. 11A-11D). The sensor system 112 may also include a sensor controller (or possibly one sensor controller per sensor if the sensor system 112 has multiple sensors), for example, to convert received / sensed sensor signals into output / measurement data that is in a format recognizable to the rest of the system.

[0013] Each sensor (e.g., each sensor "head") of the sensor system 112 is configured to scan the proximal end of the syringe 102 in a manner that outputs a measurement indicative of the depth / distance relative to that sensor. As used herein, it is understood that references to "depth" or "distance" between a sensor and a sample (e.g., the syringe 102 or a portion thereof) can refer to a distance relative to any portion of the sensor (e.g., the sensor head or the lens system of the sensor head, etc.) and can refer to a depth or distance along a straight optical path or along an optical path that has been redirected one or more times (e.g., via one or more intervening mirrors). In some embodiments, the sensor is a passive sensor, such as a confocal chromatic sensor. For example, the sensor can be a Micro-Epsilon IFS2405-30 sensor capable of detecting a depth range of about 30 mm with a sampling / measurement rate of up to about 70 kHz, and the sensor controller can be a Micro-Epsilon Confocal DT 2461 controller. Alternatively, the sensor may be another confocal chromatic sensor with sufficient precision, accuracy, sampling rate and depth range. Confocal chromatic sensors may be advantageous in that they work very well on both transparent and opaque surfaces, making them good candidates for measuring the distance of conventional flanges (e.g., glass flanges) and plungers (e.g., rubber plungers). Confocal chromatic sensors are described in further detail below with reference to FIG. 5A.

[0014] In other embodiments, each sensor of the sensor system 112 scans the proximal end of the syringe 102 not only by passing over the syringe, but also by moving closer to and / or further away from the proximal end of each syringe (i.e., shortening or lengthening the optical path between the sensor and the syringe 102, whether or not that optical path is a straight line). For example, each sensor of the sensor system 112 may be what is referred to herein as a "single focal depth" sensor. Unlike the confocal chromatic sensors described above, single focal depth sensors do not separate the light source into its different wavelength components, and therefore the sensor is associated with only a single focal length and focal depth. In these embodiments, a given sensor of the sensor system 112 is moved toward and / or away from the proximal end of the syringe until the distance with the best focus is determined, thereby providing a direct measurement of distance / depth. Single focal depth sensors are described in more detail below with reference to Figures 5B and 5C.

[0015] In yet other embodiments, the sensor system 112 may instead (or as well) include any other suitable type of sensor capable of passively or actively sensing depth / distance. For example, the sensor system 112 may include a time-of-flight (ToF) sensor (e.g., a range imaging camera) that generates / emits an optical (e.g., infrared light) signal and measures the round trip time it takes for the optical signal to reflect off a surface (e.g., a point along each of the syringes 102) and return to the sensor. For example, the ToF camera may provide 640x480 points of depth to the ToF camera with sub-millimeter granularity / precision in a single frame. The ToF camera may measure the time of flight of, for example, a laser or LED light. As another example, the sensor system 112 may include a triangulation sensor to determine depth. However, due to various factors, a triangulation sensor may not be suitable for plunger depth measurement. For example, for a syringe with a shape that precludes a large (e.g., larger ratio of plunger depth to barrel diameter) angle of incidence for triangulation, triangulation may not be possible. Furthermore, triangulation sensors typically use red or blue light that does not reflect well from the syringe glass (e.g., borosilicate glass) surface, making it very difficult to obtain a stable measurement. Still further, the flange surface of a syringe is often not perfectly flat, which further increases the difficulty of triangulation measurements. Unlike triangulation and certain other techniques, the confocal chromatic and single focal depth sensors described herein can detect borosilicate glass surfaces (and other such syringe surfaces, including plunger surfaces) with a fairly high signal-to-noise ratio.

[0016] The sensor positioning mechanism 114 of the automated inspection station 100 moves the sensors of the sensor system 112 relative to the syringe 102 according to a desired scan pattern. The automated inspection station 100 can hold each of the syringes 102 stationary as the sensor positioning mechanism 114 moves the sensors, thereby reducing the possibility of undesired syringe movement (e.g., bumping) during scanning. The sensor positioning mechanism 114 can include one or more mounting components to which the sensors of the sensor system 112 are secured, as well as one or more drive mechanisms (e.g., motors or pneumatic systems) that cause movement of these components in response to electronic control signals. In some embodiments, the sensor positioning mechanism 114 is a three-axis Cartesian robot, such as the TT-C3-4040 robot from Intelligent Actuator, Inc. Cartesian robots can be particularly well suited for embodiments in which the syringes 102 are carried in syringe tubs or similar containers, and generally have an appropriate level of precision and reliability of operation. In some embodiments, the sensor positioning mechanism 114 outputs data indicating the time and / or location (e.g., x, y coordinates) at which each measurement was captured for each of the sensors. As described below, this information can be used to identify which measurements (e.g., scan portions) correspond to which syringes 102.

[0017] In some embodiments, the automated inspection station 100 does not include a sensor positioning mechanism 114. For example, the automated inspection station 100 can hold the sensor in a fixed position above the syringes 102 as the syringe transport mechanism 110 (e.g., a star wheel or linear conveyor) passes in front of (e.g., below) each syringe.

[0018] The relative movement between the syringe 102 and the sensor caused by the sensor positioning mechanism 114 and / or the syringe transport mechanism 110 results in a particular scan pattern. For example, if a single sensor of the sensor system 112 passes over each of the syringes 102 only once (e.g., in a straight line or in a slight arc), the scan pattern is one-dimensional. As an example, a one-dimensional scan pattern is described below with reference to FIG. 6A. However, in other embodiments, the scan pattern may be two-dimensional. As an example, a two-dimensional scan pattern is described below with reference to FIG. 6B. For example, the sensor positioning mechanism 114 can cause a single sensor to perform a raster scan (i.e., multiple parallel scan lines offset by a fixed distance), a "snake" pattern (e.g., winding back and forth so that the direction does not change too abruptly at the end of each scan line), a circular scan (e.g., concentric scans with different radii), and the like. A snake pattern may be preferred over a raster pattern to avoid unnecessary re-searching by the sensor. Alternatively, if the sensor system 112 includes two or more sensors with offset positions, a single "pass" of the sensors over a given syringe 102 will result in a two-dimensional pattern (e.g., as described below with reference to Figures 8A-8C). If any fixed number of sensors is used, the optimal scan pattern may depend on the configuration of the syringe.

[0019] The automated inspection station 100 also generally includes a depth analysis unit 120 configured to process measurements generated by the sensor system 112 to determine a plunger depth of the syringe 102. The depth analysis unit 120 may include a persistent memory that stores instructions and one or more processors configured to execute the instructions to perform various operations of the depth analysis unit 120 as described herein. Alternatively, one or more of the processors of the depth analysis unit 120 may be other types of processors (e.g., an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), etc.). In various embodiments, the depth analysis unit 120 may be integrated within the same equipment as the rest of the automated inspection station 100 or may be a separate device (e.g., a laptop or desktop computer) communicatively coupled to the rest of the automated inspection station 100.

[0020] In some embodiments, the depth analysis unit 120 continuously records the output generated by the sensor system 112 as the sensor scans the syringe 102. In other embodiments, the depth analysis unit 120 does not record the output generated by the sensor system 112 and / or the sensor system 112 does not generate an output during the time interval when the sensor passes from one syringe 102 to the next syringe 102. Depending on the type of sensor, the sensor controller of the sensor system 112 may need to convert each of the sensor measurements into a distance / depth measurement in an appropriate format. For example, if the sensor includes a confocal chromatic sensor, each measurement point of the scan may be a set of amplitudes, each of which represents the amount of reflected light at a different wavelength. The sensor controller of the sensor system 112 (and / or the depth analysis unit 120) then converts each set of amplitudes into a single distance / depth measurement having an appropriate format. In other exemplary embodiments, if the sensor system 112 includes a ToF sensor, the sensor controller of the sensor detects the round trip time and converts the round trip time into a distance / depth value having an appropriate format. In embodiments where the sensor system 112 includes a single focal depth sensor, the sensor controller may generate a parameter indicative of the sharpness of the focus (e.g., the amount / intensity of reflected light) for each distance between the sensor and the syringe surface. The sensor controller and / or depth analysis unit 120 may then, for each measured parameter value, convert the parameter to a distance / depth value based on the known height of the sensor.

[0021] For one-dimensional scans, the scan can be viewed as a "depth profile." A single depth profile can reflect a set of consecutive syringes 102 (if the scan continues from one syringe to the next, e.g., within a particular syringe tub) or can reflect only a single syringe 102 (if the scan is performed discretely on each syringe). Due to the proximal viewpoint of the sensors of the sensor system 112, the smallest distance / depth will generally correspond to the flange of the syringe, while the larger distance / depth will generally correspond to the syringe plunger within the syringe barrel. The depth analysis unit 120 receives time and / or position (e.g., xy coordinates) information (e.g., from the sensor positioning mechanism 114), which enables the depth analysis unit 120 to determine which scan points (or scan lines, etc.) correspond to which syringes. Algorithms that may be employed by the depth analysis unit 120 to determine plunger depth are described in more detail below. The depth analysis unit 120 may also compare the determined plunger depths to a specified / desired range. Additionally, the depth analysis unit 120 or other software running inside or outside the automated inspection station 100 may cause the plunger depths (possibly along with an indication of whether each plunger depth is within a specified range) to be stored in local or remote memory and / or communicated to another computing system, etc.

[0022] In some embodiments, the automated inspection station 100 includes or is communicatively coupled to one or more additional components not shown in FIG 1. For example, the automated inspection station 100 may include or be coupled to a display. The display may use any suitable display technology (e.g., LED, OLED, LCD, etc.) to present information such as the plunger depth determined by the depth analysis unit 120 and / or a visual notification when the plunger depth is not within a specified range.

[0023] FIG. 2 is a perspective view of an exemplary automated inspection station 200 that may be used, for example, as the automated inspection station 100 (or part thereof) of FIG. 1. The embodiment of FIG. 2 reflects a stand-alone station, such as may be used in a laboratory (e.g., for qualification purposes), rather than a line instrument that may include multiple automated inspection stations. In the exemplary embodiment of FIG. 2, a syringe 202 (e.g., syringe 102) is held upright in a syringe tub 204 that rests on a conveyor 210 (e.g., part of a syringe transport mechanism 110). The tub 204 may be configured to hold any suitable number of syringes 202 (e.g., 36, 49, 64, 81, 100, 160, etc.). Each of the syringes 202 may be "nested" within the tub 204 to prevent excessive movement.

[0024] The conveyor 210 moves the syringe tub 204 to a position generally below the sensor 212 associated with the sensor controller 216 (e.g., collectively forming the sensor system 112), and then pauses its movement while the sensor 212 scans the syringe 202 in the syringe tub 204. FIG. 2 corresponds to an embodiment in which the sensor 212 is a confocal chromatic sensor or a single focus depth sensor, whose movement is represented in FIG. 2 by a light cone 218. In an embodiment in which the sensor 212 is a single focus depth sensor, the sensor 212 is moved towards and / or away from the syringe 202 at each measurement point to find the peak focus and thus the distance to the surface of the syringe 202. However, as mentioned above, a different type of sensor may be used instead, as long as the sensor 212 is capable of detecting the distance / depth in a non-contact / non-destructive manner in the appropriate range. The measurements of the sensor 212 are then processed by a depth analysis unit (e.g., the depth analysis unit 120), not shown in FIG. 2. The automated inspection station 200 can be used, for example, with an external computing device or system (eg, a laptop or desktop computer) or integrated processing hardware that functions as a depth analysis unit.

[0025] FIG. 3 illustrates one way in which plunger depth may be measured by the automated inspection station 100 or 200. However, it is understood that any suitable definition or technique may be used (e.g., by the depth analysis unit 120) to calculate plunger depth. In FIG. 3, an exemplary syringe 302 includes a plunger 330 disposed within a barrel 332. A proximal end of the barrel 332 (and of the syringe 302 as a whole) forms a flange 334, while a needle (not visible in FIG. 3 by a needle shield 336) is positioned at the distal end of the syringe 302. Typically, the barrel 332 and flange 334 are formed of glass, and the plunger 330 is formed of rubber. However, other materials may be used for either component (e.g., a suitable type of plastic).

[0026] In the illustrated exemplary embodiment, the plunger depth of the syringe 302 is defined as the distance between (1) the proximal (or "top") surface 340 of the flange 334 and (2) the proximal / top surface 342 of the plunger 330. However, the determination of these surfaces may be complicated by several factors. For example, the proximal flange surface 340 and / or the proximal plunger surface 342 may be uneven (e.g., undulating with distinct peaks and valleys), in which case the depth analysis unit 120 may determine the distance to the proximal flange surface 340 and / or the distance to the proximal plunger surface 342, such as by averaging the scan points corresponding to the flange 340 and the plunger 342, respectively, or by taking the peak value (minimum distance / depth) of those scan points. The depth analysis unit 120 may also ignore certain features of the flange 334 and / or the plunger 330 when determining the distance to the surfaces 340 and / or 342. 3, the plunger 330 may have a small "lug" or "dimple" 344 protruding from its proximal surface 342, in which case the depth analysis unit 120 may determine the distance to the proximal plunger surface 342 by ignoring the dimple 344 (e.g., by discarding measurements / samples that correspond to the dimple 344 before averaging). As another example, the depth analysis unit 120 may determine the distance to the proximal flange surface 340 by ignoring any chamfered edges of the flange 334 (e.g., by discarding measurements / samples that correspond to the chamfered edges before averaging).

[0027] The depth analysis unit 120 may also take into account other factors when determining distance / depth, such as the orientation of each syringe in its holder (e.g., star wheel, tub, round tray, etc.). For example, syringes in a tray or tub are typically suspended from their flanges, which may not be perfectly perpendicular to the cylindrical body (barrel) of the syringe. This may result in a slight tilt or skew, as depicted for syringe 402 (with angular displacement 430) in FIG. 4. The plunger may also be slightly skewed in the syringe barrel. In some embodiments, the depth analysis unit 120 takes into account the angular displacement of the flange and / or plunger by identifying the angle of the flange and / or plunger, and using a different algorithm to determine the plunger depth if the angle exceeds a certain threshold. For example, the depth analysis unit 120 may in such a case determine the flange surface depth based on the highest point of the proximal flange surface and determine the plunger surface depth based on the highest point of the proximal plunger surface (possibly after discarding / ignoring measurements related to dimples). Alternatively, the depth analysis unit 120 may always use the highest scanned point of the flange and / or plunger without checking if angular displacement is present. In yet other embodiments, the depth analysis unit 120 performs a mathematical transformation on the depth profile to remove or mitigate angular displacement before determining the proximal flange / plunger surface and calculating the plunger depth. Other techniques / algorithms are possible.

[0028] The filling stage is necessarily a clean process with relatively low risk of dust and debris, and the packaging environment (where plunger depth can be remeasured immediately prior to assembly of the combined device) is also typically relatively clean. However, some minimal level of dust / debris contamination is unavoidable. In some embodiments, however, the sensor system 112 is by its nature tolerant to dust, debris and other small perturbations. In particular, confocal chromatic sensors and single focus depth sensors are generally insensitive to such perturbations, and the presence of dust / debris will generally have only a minor effect on the sensor measurements.

[0029] FIG. 5A illustrates an exemplary confocal chromatic sensor 512 that may be used in the sensor system 112 of the automated inspection station 100 of FIG. 1 or another suitable automated inspection station. As seen in FIG. 5A, the confocal chromatic sensor 512 is positioned above a given syringe 502 such that the confocal chromatic sensor 512 "looks down" into the proximal end of the syringe 502. The light source 540, which may be separate from the sensor head to maintain a purely passive sensor head that does not generate heat, generates white light. The white light propagates through a fiber optic cable 542 to a fiber coupler 544 and then passes through a set of vertically aligned passive lenses 546 (e.g., six or seven lenses).

[0030] Lens 546 splits the white light into its component frequency bands (i.e., different wavelengths / colors) and focuses the light of different frequencies at different distances from sensor 512 (i.e., from the sensor head that contains lens 546). Specifically, lens 546 focuses the light at the shortest wavelength (λ min ) at the shortest distance from the sensor 512 and the longest wavelength of light (λ max) at the longest distance from the sensor 512. The amount of light reflected by a surface for a particular wavelength / frequency is a function of how well that frequency of light is focused at the point where it strikes the surface. In other words, if the surface 550 is at a particular distance within the measurement range, the intensity of the reflected light will be greatest for the particular wavelength (λ1) that is best focused at that distance of the surface 550. Thus, for any given measurement / scan point, the amount / intensity of reflected light at each wavelength indicates the depth / distance (relative to the sensor 512) at that measurement / scan point. This phenomenon is illustrated by the exemplary example plot 552 in FIG. 5A. The reflected light is analyzed by the depth analysis unit 120 after passing through a lens 546, a fiber coupler 544, and another fiber optic cable 554. As discussed above, test results show that the use of a confocal chromatic sensor is particularly well suited for depth detection of a glass syringe with a rubber plunger. Furthermore, confocal chromatic sensors can provide measurements at very high sampling rates (eg, approximately 70,000 samples per second), thereby offering high throughput compared to other conventional techniques.

[0031] The sensor controller (and / or depth analysis unit 112) then converts the intensity vs. wavelength (or intensity vs. frequency, etc.) information of each point of the scan into distance / depth, thereby generating a one-dimensional "depth profile" or a two-dimensional "depth image" (depending on the scan pattern). Exemplary depth profiles and depth images are shown in FIGS. 6A and 6B, respectively. Referring initially to FIG. 6A, depth profile 600 represents a continuous scan by a sensor of sensor system 112 (e.g., by sensor 512) over two 1 mL glass syringes positioned next to each other (e.g., inside a tub, round tray, star wheel, etc.). The scan may, for example, pass in a straight line over the syringes, or may pass in a slight arc (e.g., if the syringes are held in a rotating star wheel). In some alternative embodiments, there are one or more mirrors in the optical path between sensor 512 and syringe 502, so that sensor 512 does not have to "see" the proximal end of syringe 502 directly. In the depth profile 600, depth 620 corresponds to the proximal surface of the syringe flange, depth 622 corresponds to the proximal surface of the syringe plunger, and depth 626 corresponds to the region between consecutive syringes. The depth analysis unit 120 calculates the plunger depth 624 as the difference between the depth of surface 620 and the depth of surface 622. As described above, the depth analysis unit 120 can apply various filtering or other processing techniques, such as averaging points across each surface 620 or 622, ignoring chamfered edges and / or dimples, before calculating the plunger depth 624. The depth analysis unit 120 can determine which plunger depth corresponds to which syringe, for example, based on time (e.g., timestamp) and / or position (e.g., xy coordinates) information output by the sensor positioning mechanism 114. Although this specification is primarily described in terms of scan "points," it will be understood that in some embodiments the sensor head (e.g., a confocal chromatic sensor head) may move to measure scan lines.

[0032] 5B and 5C show an example of an alternative single focal depth sensor 560 that may be used in the sensor system 112 of the automated inspection station 100 of FIG. 1 or another suitable automated inspection station. As seen in FIG. 5B, the single focal depth sensor 560 is positioned facing (e.g., above) the proximal end of a given syringe 502, such that the single focal depth sensor 560 "looks down" into the proximal end of the syringe 502. The light 562 from the light source may be separated from the sensor head to maintain a purely passive sensor head that does not generate heat, and may be monochromatic (e.g., a single wavelength). The light 562 propagates through a lens 564. As seen in FIG. 5B, the lens 564 is associated with a single, fixed focal length, focal plane, and focal depth. Preferably, the lens 564 has a small / narrow focal depth (e.g., 1 millimeter or less). Although not shown in FIG. 5B or 5C, sensor 560 includes a camera for measuring the focus level at each distance from the syringe surface (eg, by measuring the amount / intensity of reflected light).

[0033] As the distance between the lens 564 and the syringe surface changes, the syringe surface may become more focused or less focused. FIG. 5C shows a sharpness function 582 indicative of focus, whose value may be generated by the sensor controller. The peak value of the sharpness function 582 corresponds to the focal plane shown in FIG. 5B. In particular, it is noted that the accuracy and precision of the measurement is not necessarily limited to the focal depth of the lens 564. For example, the sensor controller may calculate the position of the focal plane by determining the distance at which the focus level begins to fall off or falls below some threshold relative to the detected peak value, etc. In this way, the distance to the syringe surface may be measured with a finer granularity than the focal depth. For example, a lens 564 with a focal depth of 1 millimeter may potentially provide a measurement accuracy / granularity of 5 micrometers or less. Similar to the sensor 512 of FIG. 5A, the optical path between the sensor 560 and the syringe 502 may or may not be redirected by one or more mirrors.

[0034] The single focal depth sensor 560 has certain advantages and disadvantages over the confocal chromatic sensor 512 of FIG. 5A. The confocal chromatic sensor 512 potentially allows for a simpler mechanism to be used to move the sensor 512 relative to the syringes, since the sensor 512 does not need to be moved toward and / or away from each syringe (e.g., in the Z direction). Furthermore, since the confocal chromatic sensor 512 does not need to be moved toward or away from each syringe, the sensor 512 can capture each measurement point more quickly than the single focal depth sensor 560. Nevertheless, the single focal depth sensor 560 can potentially allow for faster scanning than the confocal chromatic sensor 512. In particular, due to both the lower cost and smaller diameter required for the cameras and lenses 564, the single focal depth sensor 560 offers the possibility of using many more sensors (e.g., 160 small, low-cost cameras / lenses for 160 syringe tabs) than would be feasible with a confocal chromatic sensor. By using multiple single focal depth sensors 560 (cameras / lenses) in parallel, in some embodiments, multiple syringes (e.g., an entire tab of syringes) can be measured almost instantly.

[0035] In some embodiments, regardless of sensor type (e.g., sensor 512 or sensor 560), a depth profile such as depth profile 600 is generated by running a single sensor of sensor system 112 over the center of each syringe 102. Running the sensor only once over each syringe 102 provides high throughput, but at the cost of reduced accuracy. In particular, accuracy may be compromised because a single pass of the sensor may not capture the highest point of the flange and / or plunger and / or account for any oblique angles of the syringe and / or plunger. However, in scenarios where lower fidelity measurements are acceptable, this may be an acceptable trade-off to achieve higher throughput.

[0036] In other embodiments, accuracy is improved by running a single sensor of sensor system 112 multiple times (e.g., 3 or 5 times) over a single syringe 102, with each successive pass offset slightly in one axis relative to the previous pass. Increasing the number of passes increases the likelihood that any subtleties in the shape or alignment of the flange and / or plunger will be captured, at the cost of reduced throughput.

[0037] In yet other embodiments, precision can be further increased by more closely scanning each syringe in two dimensions (e.g., in a raster pattern), resulting in a complete or near-complete "depth image" of the flange and / or plunger, depending on the number of scan lines and the resolution of the sensor 112. FIG. 6B shows a depth image 640 (shown here as two separate depth image portions 640-1 and 640-2) created in this manner, after software matches and interprets the individual one-dimensional raster lines. By combining the depth information from the depth image portions 640-1 and 640-2, the depth analysis unit 120 can accurately calculate the plunger depth. In other embodiments, only a single depth image is generated for each syringe 102, eliminating the need to combine separate plunger and flange depth images.

[0038] In depth image portion 640-1, the proximal plunger surface appears as a ring with dimples / lugs around the circumference of the proximal plunger surface (which depth analysis unit 120 can ignore when calculating plunger depth). In depth image portion 640-2, the outermost generally concentric portion represents the proximal flange surface. Cartesian coordinate robots are particularly well suited to quickly run a sensor over the same syringe multiple times to generate a two-dimensional scan, and a raster pattern (or snake pattern, etc.) is particularly well suited to minimizing the time for such a scan. Depth analysis unit 120 can combine one-dimensional scan lines to form a two-dimensional depth image.

[0039] Due to the relatively simple data processing required when using depth profiles (as opposed to the processing required for traditional machine vision techniques), the movement of the sensor (through the scan pattern) may limit throughput more than the data processing time. The time required to move the sensor for a complete 2D scan may be long enough to make the approach more suitable for offline measurement than for integration into a production line. However, in some embodiments, the 2D scan speed is increased by using multiple sensors (e.g., two, three or more confocal chromatic sensors or single focus depth sensors) positioned with a slight lateral offset in one axis relative to each other. As sensor technology improves and sensor size is reduced (for a given depth of field), more scan lines can be generated in parallel, further increasing the 2D scan speed.

[0040] Figure 7 is a flow diagram of an example algorithm 700 that may be implemented by the automated inspection station 100 of Figure 1. The algorithm 700 may be implemented in whole or in part by, for example, the sensor system 112, the sensor positioning mechanism 114, and the depth analysis unit 120. However, for ease of explanation, the algorithm 700 is described below with general reference to operations performed by "the system." Although the algorithm 700 is described with respect to an embodiment in which a sensor (e.g., a sensor of the sensor system 112) performs a one-dimensional scan, the same principles can be applied to a two-dimensional scan.

[0041] In the example algorithm 700, in step 702, the system waits for a trigger, such as the manual press of a button or the arrival of a syringe or syringe holder (e.g., a syringe tab). After detecting the trigger in step 704, the system performs various checks. For example, the system may check for proper operation of various hardware and / or software components, including the sensor system 112, and perform safety checks.

[0042] In step 706, the system determines whether a holder for one or more syringes is in position for scanning. For example, the system can determine whether a syringe tab has been advanced (e.g., by syringe transport mechanism 110) to an appropriate scanning position. Step 706 can be accomplished, for example, using a relatively quick pre-scan of the tab (e.g., a position check of the tab edge). In step 708, the system calibrates the sensor head at one or more positions having a known distance from the sensor head. For example, sensor positioning mechanism 114 can move a sensor of sensor system 112 to a position near syringe 102 and having a known distance from the sensor of sensor system 112. A scan controller of sensor system 112 can then determine an appropriate calibration factor (e.g., distance offset) to apply in determining the depth / distance at each scan / measurement point.

[0043] In some embodiments, the system also performs a preliminary scan (e.g., faster sensor movement and less dense scan points) to determine which portion of the tab has a syringe attached. In such embodiments, subsequent steps (e.g., step 710) can be limited to only those general locations where a syringe is present, thereby improving efficiency / throughput.

[0044] In step 710, the system moves the sensor head over the product (e.g., a single syringe or syringes in a tub, etc.) in a desired scan pattern. In some embodiments, step 710 involves moving the sensor head over each syringe at a constant speed, accelerating when it reaches an area without a syringe, and decelerating back to a constant speed when it reaches the next syringe (e.g., based on a known syringe location or a syringe location determined from a preliminary scan, etc.). As the sensor head moves to capture depth information, the system stores the measurements in a linear array in memory in step 712, as well as position data (e.g., encoder output from the sensor positioning mechanism 114). In step 714, the system processes the stored measurement data along with the position data to determine the plunger depth for each of the syringes covered by the scan. Step 714 may occur after or in parallel with the full scan in steps 710 and 712. Step 714 may include applying calibration factors, checking for any angular displacements, etc.

[0045] In step 716, the system outputs the results (e.g., the determined plunger depth with corresponding syringe indicia), which may include presenting the results on a display, storing the results in a file, communicating the results to another computing system, and / or other actions.

[0046] 8-12 show various types of automated inspection stations that may be used, for example, as automated inspection station 100 of FIG. 1, depending on the manner in which the syringes are transported.

[0047] Referring initially to FIGURE 8, an exemplary automated inspection station 800 for inspecting a syringe 802 in a tub 804 uses multiple sensors 812-1 through 812-4 (as opposed to the single sensor embodiment shown in FIGURE 2). Although FIGURE 8 shows an example in which each sensor 812 is a confocal chromatic sensor having a corresponding light cone 818, the sensors 812 may be any type of sensor described herein (e.g., a single focal depth sensor, a ToF sensor, etc.). In other embodiments, the automated inspection station 800 may include more or less than four sensors 812.

[0048] Tab 804 may be similar to tab 204 of FIG. 2, and each syringe 802 may be similar to syringe 202 of FIG. 2 and may be "nested" within tab 804 to prevent excessive movement. Although not shown in FIG. 8, tab 804 may be moved via a conveyor similar to conveyor 210. The conveyor moves syringe tab 804 to a position generally below sensors 812-1-812-4 and then pauses its movement while each of sensors 812-1-812-4 scans a different one of syringes 202 within syringe tab 804. The sensor 812 measurements are then processed by a depth analysis unit (e.g., depth analysis unit 120), not shown in FIG. 8. Automated inspection station 800 may be used, for example, with an external computing device or system (e.g., laptop or desktop computer) or integrated processing hardware that functions as the depth analysis unit.

[0049] The automated inspection station 800 can increase throughput by scanning different subsections of the tab 804 in parallel. In the depicted example, the set of sensors 812-1 through 812-4 are moved together (e.g., by the sensor positioning mechanism 114) and paused sequentially over each syringe in their respective quadrants. While positioned / paused over a given set of four respective syringes 802, the sensors 812-1 through 812-4 are moved further in smaller increments (e.g., again together by the sensor positioning mechanism 114) to scan the surfaces of the four syringes 802. For coordinated / synchronized movement, the sensors 812-1 through 812-4 can be mounted together on a common two-axis horizontal actuator.

[0050] Preferably, the shape of the tab 804 and the number of syringes 802 within the tab 804 correspond to the number of sensors 812. In the embodiment shown, for example, it may be desirable to use a tab 804 that holds a 2N by 2N array of syringes 802, where N is any integer greater than 0 (but is likely to have a larger minimum N, assuming that sensor size constrains how close the sensors 812 can be positioned relative to one another), with a spacing of N syringes 802 between each sensor 812. Each syringe 802 within the tab 804 is scanned by only a single sensor 812 (unless some level of redundancy is desired), and throughput increases as the number of sensors 812 increases (e.g., as technological improvements allow sensor size to decrease, thereby allowing more densely packed sensors 812). Mirrors and / or other optics may also be used to increase the density of the sensors 812 and therefore increase throughput. The general principles underlying the automated inspection station 800 can be extended to any size array of sensors 812 and syringes 802. For example, a 3×2 array of sensors 812 with a spacing of eight syringes 802 can be used to efficiently scan a 24×16 array of syringes 802.

[0051] 9A-9D are various perspective and overhead views of an exemplary automated inspection station 900 for inspecting syringes on a linear transport mechanism 910. As seen in FIG. 9A, the linear transport mechanism 910 provides a straight, concave channel 911A through which a chuck 904 can move (e.g., in response to a drive motor or pneumatic system, not shown in FIG. 9A). The chuck 904 is configured to hold a single syringe 902, with a portion of the syringe barrel extending through a gap 911B in the channel 911A. The chuck 904 can be a friction chuck to ensure that the syringe 902 does not rotate between scans / measurements. A high fidelity linear conveyor can move the chuck 904 (and thus the syringe 902) in small, precise increments along the channel 911A. For each increment when the chuck 904 is under the sensor 912 (e.g., a confocal chromatic sensor or other sensor type discussed herein), the linear motor scans the sensor 912 laterally (e.g., with micron-level accuracy) across the proximal end (flange) of the syringe 902 while the chuck 904 and syringe 902 remain in a fixed / stationary position. Each such scan generates a one-dimensional depth profile (e.g., similar to a portion of the depth profile 600). By varying the increment size, the measurement density and corresponding overall can be controlled. For example, a 1 mm increment size may be small enough to generate a reasonably good two-dimensional depth map for production purposes. The measurements of the sensor 912 are then processed by a depth analysis unit (e.g., depth analysis unit 120), not shown in FIG. 9. The automated inspection station 900 can be used, for example, with an external computing device or system (e.g., a laptop or desktop computer) or integrated processing hardware that functions as the depth analysis unit.

[0052] FIG. 9B provides an overhead view of scan paths 924-1 through 924-5 across syringe 902 when using automated inspection station 900. At each position of syringe 902, as syringe 902 and chuck 904 move along channel 911A, sensor 912 performs a new linear scan, depicted as the top line / arrow in FIG. 9B at each successive increment. Each syringe position shown in FIG. 9B corresponds to a different time / increment, starting from the left of FIG. 9B and proceeding to the right of FIG. 9B. In other embodiments and / or scenarios, the increments may be larger (so that scan path 924 is less than five) or smaller (so that scan path 924 is more than five).

[0053] 9C and 9D show two alternative designs that may be used for the chuck 904 of the automated inspection station 900. As seen in FIG. 9C, the first design of the chuck 904 includes two chuck halves 904-1 and 904-2 having respective halves 905-1 and 905-2 of an opening 905 (the latter not shown in FIG. 9C) and respective halves 907-1 and 907-2 of a recess 907. When the chuck halves 904-1 and 904-2 are joined around the syringe 902, the barrel of the syringe 902 protrudes through the opening 905 to ensure that the flange of the syringe nests within the recess 907. The recess 907 fits the flange precisely enough to prevent or at least substantially reduce rotation of the syringe 902 within the chuck 904. An advantage of this design is that no friction / adhesion is required to prevent rotation of the syringe 902 (although in some embodiments friction and / or adhesion can be used). A potential disadvantage is a higher risk of the glass of the syringe 902 breaking due to interaction of the glass flange with the chuck 904. This design may not be suitable if there is any significant variation in the flange dimensions (e.g., due to manufacturing tolerances of the syringe). Additionally, this design cannot prevent rotation of syringes with circular flanges.

[0054] An alternative design of the chuck 904 is shown in FIG. 9D. In this design, the chuck 904 also includes two chuck halves 904-1 and 904-2 with respective halves 905-1 and 905-2 (the latter not shown in FIG. 9D) of an opening 905, such that when the two chuck halves 904-1 and 904-2 are joined around a syringe 902, the barrel of the syringe 902 protrudes through the opening 905. However, this second design does not include a recess 907 to fix the rotational orientation of the syringe flange, and the flange remains completely outside the chuck 904. To prevent or substantially reduce rotation of the syringe, the openings 905 have a high friction and / or sticky surface to securely hold each syringe 902. In some embodiments, the chuck 904 includes one or more suction cups on the surface of the opening 905 to temporarily secure a given syringe 902 and prevent rotation or other movement relative to the chuck 904. The advantage of this second design is that there is no contact with the flange, reducing the risk of breakage. A potential disadvantage is that the high friction / sticky surface material may wear over time, becoming less effective and requiring replacement. Additionally, unless carefully designed, the surface material may leave particulates / sediment on the barrel of the syringe 902.

[0055] 10A-10D are various perspective and overhead views of another exemplary automated inspection station 1000 for inspecting a syringe 1002 on a linear transport mechanism 1010 using multiple sensors 1012. As seen in FIG. 10A, the linear transport mechanism 1010 provides a straight, concave channel 1011A through which a chuck 1004 can move (e.g., in response to a drive motor or pneumatic system, not shown in FIG. 10A). The chuck 1004 (e.g., similar to chuck 904) is configured to hold a single syringe 1002, with a portion of the syringe barrel extending through a gap 1011B in the channel 1011A. The linear conveyor can move the chuck 1004 (and thus the syringe 1002) along the channel 1011A. Unlike station 900 of FIG. 9, the linear conveyor may move chuck 1004 and syringe 1002 continuously along channel 1011A rather than in discrete increments, and the sensors 1012 may all have fixed positions relative to channel 1011A. To cover at least some range of the width of channel 1011A, automated inspection station 1000 includes five sensors 1012-1 through 1012-5 positioned along the length of channel 1011A and staggered laterally across channel 1011A with small offsets relative to each other, as seen in FIGS. 10A-10D. Sensors 1012-1 through 1012-5 may be confocal chromatic sensors (having light cones 1018-1 through 1018-5, respectively, as shown in FIGS. 10A-10C) or other sensor types discussed herein. Although five sensors 1012 are shown in FIG. 10, other embodiments may have more or fewer sensors.

[0056] The sensors 1012-1 to 1012-5 can operate one at a time (e.g., sequentially as the chuck 1004 and syringe 1002 pass under each sensor 1012) or in parallel (e.g., on different syringes 1002 held by different chucks 1004 at different positions along the channel 1011A). The movement of the chuck 1004 and syringe 1002 along the channel 1011A and the orientation of the fixed sensor 1012 cause each of the sensors 1012-1 to 1012-5 to generate a scan in a direction parallel to the length of the channel 1011A (i.e., orthogonal to the scan performed by the sensor 912 of station 900). The measurements of the sensors 1012 are then processed by a depth analysis unit (e.g., depth analysis unit 120), not shown in FIG. 10. The automated inspection station 1000 can be used, for example, with an external computing device or system (eg, a laptop or desktop computer) or integrated processing hardware that functions as a depth analysis unit.

[0057] The automated inspection station 1000 has the advantage of being fast and mechanically simple (e.g., by not requiring any of the sensors 1012 to move laterally across the width of the channel 1011A). However, the sensor size may require a relatively large spacing between the sensors 1012, which increases the risk of any given syringe 902 moving (e.g., rotating) between scans by different ones of the sensors 1012.

[0058] 11A-11D are various perspective and overhead views of an alternative automated inspection station 1100 for inspecting syringes on a linear transport mechanism 1110 using multiple sensors 1112. As seen in FIG. 11A, the linear transport mechanism 1110 provides a straight, concave channel 1111A through which a chuck 1104 can move (e.g., in response to a drive motor or pneumatic system, not shown in FIG. 11A). The chuck 1104 (e.g., similar to chuck 904) is configured to hold a single syringe 1102, with a portion of the syringe barrel extending through a gap 1111B in the channel 1111A. The linear conveyor can move the chuck 1104 (and thus the syringe 1102) along the channel 1111A. 10, the linear conveyor may move the chuck 1104 and syringe 1102 continuously along the channel 1111A rather than in discrete increments, and the sensors 1112 may all have fixed positions relative to the channel 1111A. However, unlike station 1000, the automated inspection station 1100 uses an optical system 1119 to vary the optical path between each sensor 1112 and the syringe 1102. Specifically, the optical system 1119 includes a set of mirrors 1119-1 through 1119-6, each shaped and positioned to complete an optical path between a respective one of the sensors 1112-1 through 1112-6 and the syringe 1102 being scanned. 11B and 11C are perspective and overhead views, respectively, of mirrors 1119-1 to 1119-6, and FIGS. 11A and 11D are perspective and overhead views, respectively, of automatic inspection station 1100 as a whole.

[0059] 11A-11D, mirrors 1119-1-1119-6 are angled and oriented such that each of sensors 1112 scans syringe 1102 at a different lateral offset along the width of channel 1111 A and at a different position along the length of channel 1111 A. In this manner, optical system 1119 allows sensors 1112 to scan along different portions of each syringe 1102 (as in station 1000), but without the sensor size requiring a large distance between subsequent scans by different sensors 1112 (and thus less risk of syringe 1102 rotating between scans by different sensors 1112).

[0060] The sensors 1112-1 through 1112-6 may be confocal chromatic sensors (having light cones 1118-1 through 1118-6, respectively, as shown in FIGS. 11A through 11D) or any other sensor type discussed herein. Although FIG. 11 shows six sensors 1112 and six respective mirrors 1119, other embodiments may have more or fewer sensors and mirrors. The sensors 1112-1 through 1112-6 may operate sequentially as the chuck 1104 and syringe 1102 pass under each mirror 1119-1 through 1119-6. Movement of the chuck 1104 and syringe 1102 along the channel 1111A causes each of the sensors 1112-1 through 1112-6 to generate a scan in a direction parallel to the length of the channel 1111A (i.e., orthogonal to the scan performed by the sensor 912 of station 900).

[0061] To properly orient the optical path between the sensor 1112 and the syringe 1102, each of the mirrors 1119-1 through 1119-6 may have one surface in the optical path that is angled at 45 degrees to the center / long axis of the syringe 1102 and also angled at 45 degrees to the optical axis of the respective sensor 1112. In other embodiments, other angles may be used (e.g., if any one or more of the sensors 1112 are not in a plane parallel to the plane of the flange of the syringe). For example, to further increase throughput, additional sensors may be employed that are arranged in a roughly spherical (or hemispherical) fashion around the chuck 1104, all of which are directed generally inward toward the optical system 1119. The mirrors 1119-1 through 1119-6 are preferably designed to avoid substantial degradation of or interference with the optical signal between the syringe 1102 and the sensor 1112. The optical system 1119 may include additional or alternative optics to further optimize the arrangement.

[0062] The measurements of the sensor 1112 are processed by a depth analysis unit (e.g., depth analysis unit 120) not shown in Figure 11. The automated inspection station 1100 can be used, for example, with an external computing device or system (e.g., a laptop or desktop computer) or integrated processing hardware that functions as the depth analysis unit.

[0063] 12A-12C show an alternative embodiment of an automated inspection station 1200, where multiple sensors 1212 (e.g., of a sensor system 112) are spaced around the periphery of a star wheel 1210 at different radial offsets 1222 relative to the center of the star wheel 1210. Star wheels are typically used in the filling stage, and may also be used in assembly lines for combination devices in scenarios where plunger depth measurements are relevant at that stage. While star wheels can expose the side of the syringe for image capture (i.e., machine vision techniques), some system geometries are not amenable to camera placement. In such cases, the proximal end sensing techniques described herein can provide a viable alternative.

[0064] 12A-12C show an embodiment in which there are seven confocal chromatic sensors 1212-1-1212-7 at seven different radial offsets 1222-1-1222-7. However, the automated inspection station 1200 may include more or less than seven sensors 1212, and some or all of the sensors 1212 may be of a different type (e.g., single focus depth sensors or ToF sensors). The radial offsets 1222-1-1222-7 may differ by some percentage of the width of the syringe 1202 (e.g., so that the scan lines are evenly distributed across at least the entire flange of each syringe 1202 regardless of the rotation of the flange). The rotation of the star wheel 1210 causes each scan line 1224 to arc slightly, but as long as the radius of the star wheel 1210 is large enough relative to the size of the syringe 1202, the resulting depth profile can be treated as a straight line to a first approximation. The depth analysis unit 120 can then synchronize the scans / profiles of the different sensors 1212 and calculate the plunger depth based on the synchronized profiles. The use of multiple sensors 1212 can improve the effectiveness of the system by more accurately capturing the shape / orientation of the flange and plunger without compromising the throughput of the system. In some embodiments, the star wheel 1210 uses a pneumatic-based drive system (e.g., as part of the syringe transport mechanism 110) or other mechanism (e.g., a chuck similar to chuck 904) to more securely hold the syringe in place without allowing the syringe to rotate around its central axis.

[0065] A prototype station designed according to the principles and techniques disclosed herein demonstrated high speed / throughput without sacrificing accuracy and without requiring extra handling of the syringe (e.g., without removing the syringe from the tub). In the prototype, an IFS2405-30 confocal chromatic sensor (Micro-Epsilon) was positioned over the syringe in the syringe tub using a 3-axis TT-C3-4040 Cartesian robot (Intelligent Actuator, Inc.) with a confocal DT 2461 sensor controller (Micro-Epsilon) and Visual Studio 2017 control software (Microsoft).

[0066] Gage R&R (Repeatability and Reproducibility) studies were performed to evaluate the amount of variability in the measurements provided by the prototype station. One of the metrics used was the Precision to Tolerance Ratio (P / T or PTR), defined as:

number

[0067] The prototype inspection station was largely insensitive to variations in flange orientation (i.e., variations in the amount of rotation of the asymmetric flange around the axis of the syringe barrel). FIG. 13 is a table 1300 showing the plunger depth (in mm) calculated by the prototype inspection station for 10 sample syringes over a range of different flange orientations (0 degrees to 315 degrees in 45 degree increments). As can be seen in FIG. 13, for any given sample, the measurements varied by less than 0.1 mm. Even over this full range of flange orientations, the PTR did not exceed 4.0%.

[0068] FIG. 14 is a plot 1000 comparing plunger depth determined using the confocal chromatic sensor prototype (x-axis) with plunger depth determined using a conventional optical comparator (y-axis). Both axes show the measured plunger depth in millimeters. As can be seen in FIG. 14, the prototype station reliably provided very similar results to the optical comparator. However, the prototype station provided results at a faster rate. Specifically, by capturing measurements at 70,000 samples per second (for a one-dimensional depth profile), the prototype station operated approximately 10 times faster than the optical comparator. The P-value for the confocal chromatic sensor prototype against the optical comparator was less than 0.001, indicating a high statistical similarity. There was a slight bias of 30 micrometers between the two systems, but this was statistically insignificant considering that some of the optical comparator measurements showed normal human error. With the prototype, the element of human error was largely nonexistent.

[0069] 15 is a flow diagram of an exemplary method 1500 for automatically inspecting plunger depth, or more generally, the distance between two portions of a syringe. Method 1500 can be performed by an automated inspection system, such as automated inspection station 100 of FIG. 1 or an instrument that includes automated inspection station 100.

[0070] In block 1502, the sensor system generates multiple syringe scans by scanning each of a number of syringes (e.g., syringe 102) from a proximal end perspective (e.g., while the syringes are held in an upright position by a tub, round tray, star wheel holder, or another container or mechanism, or possibly while the syringes are held or supported horizontally on their sides, etc.). Block 1502 may be performed by the sensor system 112 (e.g., sensors and sensor controller) at, for example, a filling line, a product assembly process (e.g., in-line and pre-assembly), or any other suitable process where the proximal end of each syringe is accessible for scanning. Each of the syringe scans indicates a depth (distance) relative to the sensor of the sensor system at each of a number of scan points in the scan and / or along a scan line. The depth / distance corresponds to an optical path between the sensor and the syringe, which may be a straight line or may be redirected one or more times by the optical system (e.g., by one or more mirrors such as in optical system 1119). The sensor may be, for example, an optical sensor, such as a confocal chromatic sensor (e.g., sensor 512) or a single focal depth sensor (e.g., sensor 560). Alternatively, the sensor may be another suitable sensor type, such as a ToF sensor (e.g., a ToF camera). The sensor system may generate syringe scans at a rate of at least 25,000 times per second (e.g., 70,000 times per second). The multiple syringe scans may be one or two dimensional (e.g., raster or snake scans), may be discrete scans, or may all be contained within a single continuous scan.

[0071] Block 1504 represents a process that is repeated for each of the multiple syringes (e.g., for all syringes in a tub). Block 1504 may be performed, for example, by depth analysis unit 120. In block 1504, each scan for a syringe is analyzed in block 1506 to determine two distances relative to the sensor (e.g., relative to the sensor head): a first distance to a first portion of the syringe (e.g., to a flange or to a "marker" component that protrudes inward from the inner wall of the syringe barrel) and a second distance to a second portion of the syringe (e.g., to the plunger). For example, block 1506 may include analyzing a flange region of interest (ROI) within the syringe scan to determine a first distance (e.g., by determining an average distance to the proximal surface of the flange relative to the sensor), and analyzing a plunger ROI within the syringe scan to determine a second distance (e.g., by determining an average distance to the proximal surface of the plunger relative to the sensor, possibly after ignoring / discarding samples corresponding to dimples and / or chamfered edges of the plunger).

[0072] At block 1504, a distance between the first and second portions of the syringe (e.g., the depth of the plunger) is also calculated at block 1508 based on the first and second distances determined at block 1506. Block 1508 may include, for example, subtracting the first distance from the second distance.

[0073] In some embodiments, block 1502 includes generating multiple scans using multiple sensors of a sensor system. For example, each scan may be generated using a sensor located at a different offset relative to each syringe (e.g., as shown in FIGS. 10A-10D or 12A-12C). In these embodiments, block 1506 may include determining the first and second distances by analyzing all of the sensor scans (e.g., after synchronizing the measurement data from the various scans).

[0074] In some embodiments, method 1500 includes one or more additional blocks not shown in FIG. 15. For example, method 1500 may include, prior to block 1502, using a sensor system to generate a preliminary scan of the syringe holder (e.g., a tab) and determining a location where the syringe is in the syringe holder by analyzing the preliminary scan. As another example, method 1500 may include, prior to block 1502, using the sensor system to generate one or more calibration scans at one or more locations having known distances relative to the sensor. As yet another example, method 1500 may include, after block 1508 and for each syringe, comparing the calculated distance between the first portion and the second portion to a predetermined distance range to determine if the calculated distance is within the predetermined distance range and / or storing the calculated distance in memory.

[0075] An embodiment of the present disclosure relates to a non-transitory computer-readable storage medium having computer code for performing various computer-implemented operations. The term "computer-readable storage medium" is used herein to include any medium that can store or encode a set of instructions or computer code for performing the operations, methods and techniques described herein. The medium and computer code may be of a type that is well known and available to those skilled in the art of computer software technology, or may be specially designed and constructed for the purposes of the embodiments of the present disclosure. Examples of computer-readable storage media include, but are not limited to, magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and holographic devices, magneto-optical media such as optical disks, and hardware devices specially configured to store and execute program code, such as ASICs, programmable logic devices ("PLDs"), and ROM and RAM devices.

[0076] Examples of computer code include machine code, such as that produced by a compiler, and files containing relatively high-level code executed by a computer using an interpreter or compiler. For example, an embodiment of the present disclosure may be implemented using Java, C++, or other object-oriented programming languages ​​and development tools. Further examples of computer code include encryption and compression code. Furthermore, an embodiment of the present disclosure may be downloaded as a computer program product, which may be transferred from a remote computer (e.g., a server computer) to a requesting computer (e.g., a client computer or a different server computer) over a transmission channel. Another embodiment of the present disclosure may be implemented in hardwired circuitry in place of, or in combination with, machine-executable software instructions.

[0077] As used herein, the singular terms "a," "an," and "the" can include plural references unless the context clearly indicates otherwise.

[0078] As used herein, the terms "connect," "connected," and "connection" refer to an operative coupling or linking (and connections depicted in the drawings represent). Connected components may be directly or indirectly coupled to one another, for example, through a set of separate components.

[0079] As used herein, the terms "nearly", "substantially", "substantial" and "about" are used to describe and explain small variations. When used in conjunction with an event or circumstance, these terms may refer to not only the exact occurrence of the event or circumstance, but also to the occurrence of an event or circumstance that is very close. For example, when used in conjunction with a numerical value, these terms may refer to a variation of the numerical value of ±10% or less, such as ±5% or less, ±4% or less, ±3% or less, ±2% or less, ±1% or less, ±0.5% or less, ±0.1% or less, or ±0.05% or less. For example, two numerical values ​​can be considered "substantially" identical if the difference between the two numerical values ​​is ±10% or less of the average of the numerical values, such as ±5% or less, ±4% or less, ±3% or less, ±2% or less, ±1% or less, ±0.5% or less, ±0.1% or less, or ±0.05% or less.

[0080] Additionally, amounts, ratios, and other numerical values ​​may be presented herein in a range format, with the understanding that such range format is used for convenience and brevity and includes numerical values ​​expressly specified as the limits of the range, but should be understood in a flexible manner to include all individual numerical values ​​or subranges subsumed within the range, as if each numerical value and subrange were expressly specified.

[0081] Although the present disclosure has been described and illustrated with reference to specific embodiments thereof, these descriptions and illustrations are not intended to limit the present disclosure. Those skilled in the art will understand that various modifications may be made and equivalents may be substituted without departing from the true spirit and scope of the present disclosure as defined by the appended claims. The illustrations may not necessarily be to scale. Due to manufacturing processes, tolerances, and / or other reasons, there may be differences between the technical expressions in the present disclosure and the actual devices. There may be other embodiments of the present disclosure that are not specifically illustrated. The present specification and drawings (other than those in the claims) should be regarded as illustrative rather than restrictive. Changes may be made to adapt a particular situation, material, composition of matter, technique, or process to the objective, spirit, and scope of the present disclosure. All such modifications are intended to be within the scope of the claims appended hereto. Although the techniques disclosed herein have been described with reference to particular operations being performed in a particular order, it will be understood that these operations may be combined, sub-divided, or re-ordered to form equivalent techniques without departing from the teachings of this disclosure. Thus, unless specifically indicated herein, the order and grouping of operations is not intended to be limiting of the disclosure.

Claims

1. 1. A method for automated inspection of a syringe, comprising: generating a plurality of syringe scans by scanning each of a plurality of syringes with a sensor system including a sensor, each of the plurality of syringe scans indicating a distance relative to the sensor; For each of the plurality of syringes, analyzing, with one or more processors, each of the syringe scans to determine (i) a first distance from the sensor to a first portion of the syringe, and (ii) a second distance from the sensor to a second portion of the syringe; calculating, by the one or more processors, a distance between the first portion and the second portion based on the first distance and the second distance; The method includes:

2. The method of claim 1, wherein the sensor is an optical sensor or a time-of-flight (ToF) sensor, and optionally, the optical sensor is a confocal chromatic sensor or a monofocal depth sensor.

3. The method of claim 2 , wherein the first portion of the syringe is a flange of the syringe and the second portion of the syringe is a plunger of the syringe.

4. Analyzing each syringe scan includes: analyzing a flange region of interest (ROI) within each of the syringe scans to determine the first distance; analyzing a plunger ROI within each of the syringe scans to determine the second distance; The method of claim 3 , comprising:

5. 2. The method of claim 1, wherein generating the plurality of syringe scans comprises generating a continuous scan including the plurality of syringe scans, or generating the plurality of scans of the plurality of syringes at a measurement rate of at least 25,000 measurements per second.

6. The method of claim 1 , wherein generating a plurality of syringe scans comprises generating a plurality of one-dimensional scans or a plurality of two-dimensional scans.

7. The method of claim 1 , further comprising moving the sensor relative to the plurality of syringes while generating the plurality of syringe scans.

8. the sensor is a single focus depth sensor; and 8. The method of claim 7, wherein moving the sensor relative to the plurality of syringes additionally comprises moving the single focal depth sensor toward and / or away from each syringe of the plurality of syringes while generating the plurality of syringe scans.

9. The method of claim 1 , wherein generating the plurality of syringe scans comprises scanning each of the plurality of syringes while the plurality of syringes is held in an upright position by a syringe holder.

10. generating a preliminary scan of the syringe holder with the sensor system prior to generating the plurality of syringe scans; determining locations of the plurality of syringes within the syringe holder by analyzing the preliminary scan; The method of claim 9 further comprising:

11. The method of claim 1 , further comprising generating, with the sensor system, one or more calibration scans at one or more positions having a known distance relative to the sensor prior to generating the plurality of syringe scans.

12. 2. The method of claim 1, further comprising: comparing, by the one or more processors, the calculated distance between the first portion and the second portion to a predetermined distance range to determine whether the calculated distance is within the predetermined distance range.

13. The method of claim 1 , further comprising storing, by the one or more processors, the calculated distance in a memory for each of the plurality of syringes.

14. 1. An automated inspection system, comprising: a sensor system comprising a sensor and configured to generate a plurality of syringe scans by scanning each of a plurality of syringes, each of the plurality of syringe scans indicating a distance relative to the sensor; One or more processors configured to perform the method according to any one of claims 1 to 13; Automated inspection system including:

15. One or more tangible, non-transitory computer readable media storing instructions that, when executed by one or more processors, cause the one or more processors to perform the method of any one of claims 1 to 13.

Images