Treadmill control system

Abstract

A treadmill control system (200, figure 2) for use with a treadmill 100 comprising a camera (204, figure 2) to obtain image data of a user on the treadmill and a controller (206, figure 2) to receive

Description

FIELD The present invention relates to a treadmill control system, and in particular but not exclusively to a treadmill control system for detecting falls. BACKGROUND Treadmills are widely used by people both at home and in public gyms, allowing them to improve health and fitness by running without requiring an open space. However, treadmills are large pieces of equipment which can be dangerous for people if used incorrectly or unsafely. In particular, if a person falls while using a treadmill, that can cause injuries due to impact with a moving running surface which can subsequently propel the person into other objects or surfaces after falling. Typically treadmills are provided with a safety cord with a clip to be attached to the person using the treadmill (for example, to an item of clothing worn by the person). If the safety cord is used properly, the cord will extend to its full length as the user falls. Once full extension of the safety cord is reach, an emergency stop is activated to stop movement of the running surface. However, the initial set up of the safety cord requiring direct action by the user means such safety cords are rarely utilised by people using treadmills, and often utilised incorrectly when they are used (for example, people rarely adjust the length of the safety cord to be appropriate for their height). In practice, therefore, safety cords provide little protection to people using treadmills. The present invention has been devised with the foregoing in mind. SUMMARY According to a first aspect there is provided a treadmill control system for use with a treadmill. The system may comprise a camera configured to obtain image data of a user on a treadmill. The system may also comprise a controller configured to receive the image data from the camera. The controller may comprise a machine learning algorithm trained to determine falling of the user from the image data. The controller may be further configured to generate, on determining falling of the user, a control signal configured to stop movement of a running surface of the treadmill. The control system of the present invention may provide an automatic safety provision for a user of a treadmill which can be utilised with no action required by the user. The control system may remove the need to manually attach a conventional safety cord or manually activate another safety mechanism. That may also improve accuracy of fall detection compared to sensor-based approaches, which may not differentiate between sensor readings resulting from actions of the user on the treadmill and sensor readings due to actions or presence of other persons or objects near the treadmill. Instead, the control system detects falling based on actual image data of a user on the treadmill, which may reduce false positive detection of falling. In addition, using image data of the user on the treadmill may improve speed of fall detection and in turn response speed of the control system to stop movement of the running surface following detection of a user falling. By using image data, the control system may detect falling without requiring the user to have already reached a certain location on the treadmill to trigger a safety mechanism (for example, travelled sufficiently rearward for a safety cord to reach full extension and activate an emergency stop), which can delay the running surface from being stopped. Detecting falling of a user on the treadmill based on image data may also enable the control system to detect falling in its early stages, before the user has actually fallen. Improved speed of fall detection and response speed of the control system to stop movement of the running surface may reduce or remove a risk of the user landing on a moving running surface on falling, which may reduce a likelihood or severity of injury sustained on falling (for example due to impact, or friction burns from landing on a moving running surface). That may also reduce a risk of the user being propelled off the treadmill by the moving running surface into nearby objects or surfaces after falling, which can often exacerbate an initial injury sustained on falling or even cause additional injuries. That may also reduce contact by the user with such safety mechanisms on the treadmill other than the running surface, reducing a spread of common health conditions and illnesses spread by contact (which are common in public gyms with multiple users sharing equipment). That may be particularly advantageous with respect to a safety cord which is typically more difficult to wipe over than other surfaces of a treadmill. The system may further comprise means for detecting a position of the user on the treadmill. The means may be or comprise a non-contact sensor. The controller may be configured to receive data indicative of a position of the user on the treadmill from the means for detecting a position of the user on the treadmill. The controller may also be configured to generate, on determining the position of the user on the treadmill is within a pre-determined distance of a selected location on the treadmill, a control signal configured to stop movement of the running surface of the treadmill. That may provide redundancy or backup to the image data based approach of the control system to further improve safety, whilst maintaining an automated safety provision requiring no manual interaction or initiation by the user. The selected location on the treadmill may be at or adjacent a rearward end of the running surface of the treadmill. The controller may be configured to generate a control signal configured to stop movement of the running surface of the treadmill on determining both falling of the user and a position of the user on the treadmill being within a pre-determined distance of the selected location on the treadmill. That may reduce or minimize a risk of false positive fall detection, for example if image data of the user on the treadmill shows the user stumbling but not actually falling, or if the means for detecting a position of the user on the treadmill detects actions or presence of other persons or objects near the treadmill rather than the user itself. That may avoid or reduce a risk of movement of the running surface being inadvertently stopped which might in fact cause a user to fall. The system may comprise a sensor configured to detect performance of one or more gestures by the user on the treadmill. The controller may be configured to receive data from the sensor. The controller may also be configured to determine performance of one or more pre-determined gestures by the user on the treadmill from the sensor data. The controller may be further configured to generate, on determining performance of a pre-determined gesture by the user on the treadmill, a control signal to control operation of the treadmill. That may enable the treadmill to be operated in an essentially contactless manner by a user, enabling both safety provision and operational control of the treadmill by the user without manual or physical contact between the user and the treadmill. That may in turn further reduce spread of common health condition spread by contact. That may also allow the treadmill to be provided without a manual user interface or console (for example, comprising buttons or switches including touch-sensitive buttons, or a touch screen display) via which a user can control treadmill operation. Such interfaces are typically located adjacent a front end of the treadmill for use by a user during running. However, it is typically difficult for users to use such manual interfaces easily or accurately whilst moving on the treadmill, particularly at high speeds. A contactless treadmill which a user may control using pre-determined gestures may simplify use of the treadmill whilst the user is running, improving ease and convenience. Removal of a manual interface or console may also improve display capabilities of the treadmill, for example, by enabling a significantly larger display screen to be provided in front of a user on the treadmill, the view of which is not then obstructed by the manual interface or console. That may improve an immersive experience for the user on the treadmill. In addition, removal of a manual interface or console may reduce or eliminate protruding structures on the treadmill which a user may otherwise impact during a fall, which may further improve safety for the user. Removing a manual user interface or console for controlling treadmill operation which multiple users contact directly may also reduce or remove a need for cleaning of the treadmill. Continuous or prolonged use of liquid cleaning solutions such as antibacterial solutions and / or antiviral solutions can cause electronic systems in manual user interfaces to malfunction. Removing the manual user interface and in turn reducing exposure of the treadmill to such liquid cleaning solutions may reduce or prevent damage to electronic systems of the treadmill, thereby increasing a working life of the treadmill. The sensor configured to detect performance of one or more gestures by the user on the treadmill may be or comprise the camera configured to obtain image data of the user on the treadmill. The controller may comprise a machine learning algorithm trained to determine performance of the one or more pre-determined gestures by the user on the treadmill from the sensor data. The controller may comprise a machine learning algorithm trained to determine falling of the user from the image data, and determine performance of the one or more pre-determined gestures by the user on the treadmill from the image data. The controller may be configured to generate a control signal to control one or more of a speed of the running surface of the treadmill, and an inclination of the running surface of the treadmill. The means for detecting a position of the user on the treadmill may comprise a non-contact sensor. The non-contact sensor may be or comprise one or more of a range sensor configured to provide data indicative of a distance between the user and the range sensor, and a location or boundary sensor configured to provide data indicative of whether the user has reached or is within a pre-determined distance of the sensor. The system may not comprise a manual user interface for receiving user instructions. According to a second aspect there is provided a treadmill comprising the treadmill control system of the first aspect. The treadmill may comprise a display screen configured to display a virtual environment through which the user on the treadmill can navigate. A structure of the treadmill at or adjacent a forward end of the running surface may consist essentially of a display screen. According to a third aspect there is provided a method of controlling a treadmill. The method may comprise obtaining image data of a user on a treadmill. The method may further comprise providing the obtained image data to a machine learning algorithm trained to determine falling of the user from the image data. The method may further comprise stopping movement of a running surface of the treadmill if falling of a user is determined by the machine learning algorithm. The method may be a computer-implemented method. The method of the first aspect may be performed by the control system of the first aspect and / or by the treadmill of the second aspect. Features which are described in the context of separate aspects and embodiments of the invention may be used together and / or be interchangeable wherever possible. Similarly, where features are described in the context of a single embodiment for brevity, those features may also be provided separately or in any suitable sub-combination. BRIEF DESCRIPTION OF DRAWINGS Embodiments of the invention will now be described by way of example only with reference to the following drawings in which: FIG. 1 shows a treadmill according to an embodiment of the present invention; FIG. 2 shows a control system for a treadmill according to an embodiment of the present invention; FIG. 3 shows an alternative control system for a treadmill according to another embodiment of the present invention, comprising an additional sensor for detecting a position of a user on the treadmill; FIGs. 3A and 3B show a treadmill incorporating the control system shown in FIG. 3; FIG. 4 shows another alternative control system for a treadmill according to another embodiment of the present invention, comprising a controller configured to recognise performance of gestures by a user on the treadmill to control operation of the treadmill; and FIGs. 4A to 4C show examples of gestures which may be performed by a user on the treadmill to control operation of the treadmill; and FIG. 5 shows a method for controlling a treadmill according to an embodiment of the present invention. Like reference numerals in different Figures may represent like elements. DETAILED DESCRIPTION Figure 1 shows a treadmill 100. The treadmill 100 comprises a running surface 102 on which a user of the treadmill 100 can walk, run or jog in use. In the embodiment shown, the running surface 102 comprises a belt 102a provided on one or more rotatable elements (for example, rollers, not shown) to enable the belt 102a to move continuously underneath the feet of the user as the user moves on the treadmill 100 in a conventional manner. In the embodiment shown, the rotatable elements are actively rotated, either directly or indirectly, by a motor (not shown) connected to a power source, in order to drive movement of the running surface 102. However, it will be appreciated the rotating elements may be passively rotated, for example rotated in response to movement of the belt 102a caused by movement of the user on the treadmill 100. It will also be appreciated the treadmill 100 may comprise any suitable running surface 102 comprising an active or passive movable element to enable the running surface 102 to move 7 continuously underneath the feet of the user as the user moves on the treadmill 100, for example a unidirectional running surface or an omnidirectional running surface. Figure 2 shows an embodiment of a control system 200 for use with a treadmill such as the treadmill 100 described with respect to Figure 1. The control system 200 comprises a camera 204. The camera 204 is configured to obtain image data of a user on the treadmill 100 as the user moves on the running surface 102. The control system 200 also comprises a controller 206. The controller 206 is configured to receive the image data obtained by the camera 204. The controller 206 comprises a machine learning algorithm trained to determine falling of the user on the treadmill 100 using the image data of the user on the treadmill 100 obtained by the camera 204. If the controller 206 determines falling of the user on the treadmill 100, the controller 206 is configured to generate a control signal to stop or prevent movement of the running surface 102 of the treadmill 100. In the embodiment shown, one or more components of the control system 200 are provided separately from the treadmill 100 itself such that the control system 200 can be retrofitted or installed on any treadmill 100. The camera 204 is structurally separate from the treadmill 100 such that a position of the camera 204 can be selected or adjusted as desired, although that is not essential and the camera 204 may be integral to the treadmill 100 and provided in a substantially fixed position on the treadmill 100. However, it will be appreciated one or more components of the control system 200 may be integral to the treadmill 100. The camera 204 may be positioned in any suitable location to capture image data of the user on the treadmill 100, for example substantially in front of a user on the treadmill 100, substantially to a lateral side of the user on the treadmill 100, or substantially behind a user on the treadmill 100 (for example, relative to an expected or intended orientation of the user on the treadmill 100 in use). In the embodiment shown, the controller 206 comprises a transceiver and is in wireless communication with the camera 204 (for example, via Wi-Fi®, Bluetooth® or any suitable wireless communication protocol such as infrared communication), although that is not essential and the controller 206 may be in wired communication with the camera 204. The controller 206 may be integral to the treadmill 100, or may be provided separately or remotely from the treadmill 100 (for example, on a remote server) and be in wired or wireless communication with the treadmill 100. The machine learning algorithm implemented by the controller 206 may be or comprise a convolutional neural network which are well-suited to analysing image data, although any suitable machine learning algorithm may alternatively be used. In the embodiment shown, the running surface 102 is actively driven by a motor as described above. The control signal generated by the controller 206 causes rotation of the motor to cease 8 via direct electronic or electrical control, in turn preventing movement of the running surface 102 to which the motor is connected. Additionally or alternatively, the control signal generated by the controller 206 may cause actuation of a detent or blocking element or mechanism configured to physically prevent rotation of the motor, or may cause the running surface 102 to become disconnected from the motor, in turn preventing movement of the running surface 102 to which the motor is connected. For a running surface 102 provided on passively driven rotatable elements, the control signal generated by the controller 206 may cause actuation of a detent or blocking element configured to physically prevent rotation of the rotatable element to which the running surface 102 is connected, or may cause the running surface 102 to become disconnected from the rotatable element, in turn preventing movement of the running surface 102. Figure 3 shows another embodiment of a control system 300 for use with a treadmill such as the treadmill 100 described with respect to Figure 1. The control system 300 is similar to the control system 200 described with respect to Figure 2, with like reference numerals indicating like elements. The control system 300 comprises a camera 304 and a controller 306 substantially as described above with respect to the control system 200 shown in Figure 2. The control system 300 further comprises a sensor 308 for detecting a position of the user on the treadmill 100. In the embodiment shown, the sensor 308 comprises an ultrasonic range sensor, comprising a transmitter to emit an ultrasonic signal and a receiver configured to receive a reflected ultrasonic signal, in order to determine a distance of the user from the sensor 308 (as illustrated in Figure 3A). However, the sensor may additionally or alternatively comprise any suitable range sensor, for example an infrared sensor, or a LIDAR sensor. The sensor may additionally or alternatively comprise one or more weight, force or pressure sensors provided underneath the running surface 102 to detect a position of the user on the treadmill 100. Additionally or alternatively, the sensor may comprise a boundary sensor 308’, for example a laser beam or light curtain sensor provided on the treadmill 100 (as illustrated in Figure 3B). The light curtain or beam sensor 308’ comprises a transmitter 308a’ configured to emit a light beam and a receiver 308’b configured to receive the light beam emitted by the transmitter 308’a. In the embodiment shown, the transmitter 308a’ and the receiver 308’b are disposed on respective hand rails located on opposing lateral sides of the treadmill 100, although that is not essential and the transmitter 308a’ and the receiver 308b’ may be provided on opposing lateral sides of the treadmill 100 at any suitable location. If the beam between the transmitter 308’a and the receiver 308’b is interrupted, that may indicate the user has reached the location of the sensor 308’ on the treadmill 100. The controller 306 is configured to receive data indicative of a position of the user on the treadmill 100 from the sensor 308, 308’. In the embodiment shown, the controller 306 is in wireless communication with the sensor 308, 308’ (for example, via Wi-Fi®, Bluetooth® or any suitable wireless communication protocol such as infrared communication), although that is not essential and the controller 306 may be in wired communication with the sensor 308, 308’. The controller 306 is configured to determine, based on the data received from sensor 308, 308’, whether the position of the user on the treadmill 100 is at or within a pre-determined distance of a selected location on the treadmill 100. In the embodiment shown, the selected location is substantially the rearward end of the running surface 102 of the treadmill 100, although that is not essential and any suitable location may alternatively be used. If the controller 306 determines the user is at or within a pre-determined distance of the selected location on the treadmill 100, the controller 206 is configured to generate a control signal to stop or prevent movement of the running surface 102 of the treadmill 100. Alternatively, the controller 306 may be configured to generate a control signal to stop movement of the running surface 102 of the treadmill 100 only on determining both falling of the user (based on image data received from the camera 304) and a position of the user on the treadmill 100 being at or within a pre-determined distance of the selected location on the treadmill 100 (based on data received from the sensor 308, 308’). If only one of the two determinations is made, the controller 206 may not generate a control signal to stop movement of the running surface 102 of the treadmill 100. In alternative arrangements, the control system 300 may comprise a plurality of different sensors 308, 308’ configured to detect a position of the user on the treadmill 100. For example, the control system 300 may comprise both a range sensor 308 configured to provide data indicative of a distance between the user and the range sensor 308, and a location or boundary sensor 308’ configured to provide data indicative of whether the user is within a pre-determined distance of (for example, at the location of) the boundary sensor 308’. The controller 306 may be configured to receive data from each of the different sensors 308, 308’ (for example, the range sensor 308 and the boundary sensor 308’). The controller 306 may be configured to independently determine whether the data from each respective sensor 308, 308’ indicates a position of the user on the treadmill 100 is within a pre-determined distance of a selected location on the treadmill 100 (for example, the rearward end of the running surface 102 of the treadmill 100). The controller 306 may be configured to generate a control signal to stop movement of the running surface 102 of the treadmill 100 only on determining the data from each of the plurality of sensors indicates a position of the user on the treadmill is within a pre-determined distance of the selected location on the treadmill 100. That may avoid false positive detection of a user falling based on data from a single sensor which may detect actions or presence of persons or objects other than the user on the treadmill 100. For example, a boundary sensor 308’ located at or adjacent the rearward end of the running surface 102 of the treadmill 100 may be inadvertently activated by a person, animal or object passing near the rearward end of the running surface 102. If the controller 306 were to generate a control signal to stop movement of the running surface 102 based on data from the boundary sensor 308’ alone, the running surface 102 may be erroneously stopped when the user has not fallen or is not likely to fall. Rather, stopping movement of the running surface 102 based on that data alone could cause the running surface 102 to stop unexpectedly for the user, which could in fact increase the risk of the user falling and becoming injured. By cross-referencing data from multiple different sensor types 308, 308’ regarding a position of the user on the treadmill 100 relative to a selected location on the treadmill 100 (for example, the rearward end of the running surface 102 of the treadmill 100), erroneous stopping of the movement of the running surface 102 may be reduced or eliminated, improving accuracy of fall detection and in turn improving safety for a user on the treadmill 100. That may also reduce or avoid the need for conventional safety cords to which a user can attach themselves, which are often ineffective as users rarely adjust the safety cord to the required length (meaning a user can be thrown from the treadmill 100 before the safety cord causes movement of the running surface 102 to stop). It will also be appreciated data from one or more of the sensors 308, 308’ described above may be utilised to implement staggered speed control based on a position of the user on the treadmill 100 relative to the selected location on the treadmill 100. A plurality of distance thresholds relative to the selected location on the treadmill 100 may be set. For example, first, second and third distance thresholds may be set, with the first threshold being the largest (i.e., furthest distance from the selected location on the treadmill 100) and the third threshold being the smallest (i.e., closest distance from the selected location on the treadmill 100, for example at the selected location on the treadmill 100). Using the example of an ultrasonic range sensor 308, if data from the ultrasonic range sensor 308 shows the user is below the first distance threshold, the controller 306 may be configured to generate a control signal to cause a warning to be provided to the user without adjusting a speed of movement of the running surface 102. For example, the control signal may cause a visual warning to be provided on a display or using one or more light sources (such as LEDs provided on the treadmill 100), or cause an audio warning to be emitted by a speaker or buzzer. If data from the ultrasonic range sensor 308 shows the user is below the second distance threshold, the controller 306 may be configured to generate control signals to both cause a warning to be provided to a user and reduce a speed of movement of the running surface 102. If data from the ultrasonic range sensor 308 shows the user is below the third distance threshold, the controller 306 may be configured to generate a control signal to stop movement of the running surface 102. It will also be appreciate a similar approach may be implemented using a plurality of boundary sensors (such as laser beams 308’) located at positions on the treadmill 100 corresponding to the respective first, second and third distance thresholds. Figure 4 shows another embodiment of a control system 400 for use with a treadmill such as the treadmill 100 described with respect to Figure 1. The control system 300 is similar to the control systems 200, 300 described with respect to Figures 2 and 3, with like reference numerals indicating like elements. The control system 400 comprises a camera 404 and a controller 406 substantially as described above with respect to the control system 200 shown in Figure 2. In the embodiment shown, the controller 406 is configured to determine or recognise performance of one or more predetermined gestures by the user on the treadmill 100 from the image data obtained by the camera 404. If the controller 406 determines performance of a pre-determined gesture by the user on the treadmill 100, the controller 406 is configured to generate a control signal to control operation of the treadmill 100. In the embodiment shown, the controller 406 is configured to recognise a plurality of predetermined gestures performed by a user with their hand and / or arm whilst on the treadmill 100. For example, Figure 4A shows a first gesture comprising the user forming a substantially closed fist facing frontwards (Figure 4A1), and a second gesture comprising the user forming a substantially open palm facing frontwards (Figure 4A2). The user performing the first gesture and second gesture in succession, in that order, is recognised by the controller 406 as a command to initiate movement of the running surface 102 of the treadmill 100. The user performing the second gesture and the first gesture in succession, in that order, is recognised by the controller 406 as a command to stop movement of the running surface 102 of the treadmill 100. Figure 4B shows a gesture comprising the user raising or lowering the forearm in front of the body by rotating around the elbow. The user performing the gesture and raising the forearm upwards (along the direction of arrow 4B1) is recognised by the controller 406 as a command to increase an incline angle of the running surface 102 of the treadmill 100. The user performing the gesture and lowering the forearm downwards (along the direction of arrow 4B2) is recognised by the controller 406 as a command to decrease an incline angle of the running surface 102 of the treadmill 100 (including orienting the running surface 102 at a decline angle). Figure 4C shows a gesture comprising the user forming a substantially open palm oriented with the palm facing substantially upwards (Figure 4C1) or substantially downwards (Figure 4C2), and curling the fingers in towards the surface of the palm. The user performing the gesture with the palm facing 12 substantially upwards is recognised by the controller 406 as a command to increase a speed of movement of the running surface 102 of the treadmill 100. The user performing the gesture with the palm facing substantially downwards is recognised by the controller 406 as a command to decrease a speed of movement of the running surface 102 of the treadmill 100. Upon recognising or determining one of the pre-determined gestures, the controller 406 is configured to generate a respective control signal to control operation of the treadmill 100 according to the recognised or determined gesture. It will also be appreciated any suitable pre-determined gesture(s) may be employed by the user to control operation of the treadmill 100. For example, gestures may be performed by one or both hands, one or more digits (for example, fingers and / or thumbs) and / or one or both arms of the user, and / or by head or eye movement of a user (such as by monitoring or tracking eye movement of the user using image data obtained by the camera 404). The pre-determined gestures may be configured to control operation of the treadmill 100 other than movement, speed or inclination of the running surface 102. For example, the controller 406 may be configured to recognise or determine a gesture which causes a cup or bottle holder to be moved from a stowed position (for example, located within or on or otherwise integrated into a side rail 103 of the treadmill 100) to an operating position configured to receive a cup or bottle of the user. The controller 406 may be configured to recognise or determine a gesture which mimics the motion of a user raising a cup or bottle to their mouth in order to cause the cup or bottle holder to be moved to the operating position. A cup or bottle holder integrated into a side rail 103 of the treadmill may avoid liquid containers being positioned adjacent a main user interface or console of the treadmill 100 (as is typical on conventional treadmills), reducing a risk of liquid being spilled or tipped onto electronic systems of the treadmill 100 (particularly when a user runs at high speed which can cause instability). Additionally or alternatively, the controller 406 may be configured to recognise or determine one or more gestures to control a visual or audiovisual output of the treadmill 100, for example by controlling a display screen and / or one or more light sources and / or speakers on the treadmill 100. In the embodiment shown, the controller 406 comprises a second machine learning algorithm trained to determine or recognise a gesture performed by the user on the treadmill 100. The second machine learning algorithm implemented by the controller 206 may be or comprise a convolutional neural network which are well-suited to analysing image data, although any suitable machine learning algorithm may alternatively be used. It will also be appreciated a single machine learning algorithm may be used to both determine falling of the user and recognise one or more pre-determined gestures performed by the user on the treadmill 100, based on image data obtained from the camera 404. Alternatively or additionally, one or more other sensors may be provided to detect performance of a gesture by the user on the treadmill. For example, an infrared sensor or ultrasonic sensor may be used and the sensor data processed to capture the spatial position and motion of the user in order to determine performance of one or more pre-determined gestures. The controller 406 being configured to determine both falling of a user on the treadmill 100 and performance of one or more pre-determined gestures by the user on the treadmill 100 may enable the treadmill 100 to be provided without a manual user input interface or console (as is typically provided adjacent a forward end of the treadmill on conventional treadmills), and the user to both use and operate the treadmill 100 in a contactless manner (without requiring any manual physical contact between the user and the treadmill 100). In addition or alternatively to the controller 406 being configured to determine or recognise performance of a gesture or gesture sequence by the user on the treadmill 100 to stop movement of the running surface 102, the controller may be configured to generate a control signal to stop movement of the running surface 102 based on data received from one or more sensors indicative that a user is no longer present on the treadmill 100. For example, if data from a weight, force or pressure sensor indicates there is no user on the treadmill 100, or if data from a range sensor such as an ultrasonic range sensor 308 indicates there is no user on the treadmill 100, or if image data from the camera 404 indicates there is no user on the treadmill 100, the controller 406 may generate a control signal to stop movement of the running surface 102. That may improve safety by preventing movement of the running surface 102 when a user is not on the treadmill 100. That may reduce or prevent instances of a user unwittingly stepping onto a moving running surface 102 and becoming injured. That may also enable the treadmill 100 to avoid wasted energy usage when no user is present, and / or may enable the treadmill 100 (or at least the controller 406) to enter a sleep mode in order to save further energy. Likewise, if data from one or more sensors (for example a weight, force or pressure sensor, ultrasonic range sensor 308, camera 404, beam sensor 308’, or a motion sensor) indicates a user has gotten on to the treadmill 100, the controller 406 may exit the sleep mode and generate a control signal to initiate movement of the running surface 102. By removing the need for any manual user input interface or console at or adjacent a forward end of the treadmill 100, a structure of the treadmill 100 at or adjacent a forward end of the running surface 102 may consist essentially of a display screen 101 as shown in Figures 1, 3A and 3B. The display screen 101 may be viewed without obstruction by any objects or structures located in front of the user on the treadmill 100 in use, substantially improving display capabilities of the treadmill 100 compared to conventional treadmill displays (for example, by being able to provide a substantially larger display screen without obstruction). That may be particularly advantageous, for example, to display a virtual environment through which a user can navigate whilst using the treadmill 100. Motion of the running surface 102 of the treadmill 100 (for example, a speed of the belt 102a of the running surface 102) may be monitored, and the virtual environment displayed on the display screen may be adjusted in accordance with motion of the running surface 102 (for example, to display travel through the virtual environment at a speed corresponding to a speed of the running surface 102). That way, the display of the virtual environment will be synchronised to the speed of the running surface 102. In addition, an orientation and / or incline angle of the running surface 102 may be adjusted to reflect a change of orientation (direction of travel through the virtual environment, for example an azimuthal angle) or incline angle in the virtual environment displayed on the display screen, to provide an immersive experience for the user on the treadmill 100. The controller 206, 306, 406 may be configured to receive information regarding the orientation and / or incline angle of the virtual environment and may generate a control signal to adjust an orientation and / or incline angle of the running surface 102 accordingly. Similarly, the controller 206, 306, 406 may be configured to generate, based on motion of the running surface 102, a control signal to control display of a virtual environment on a display screen of the treadmill 100. In some arrangements, the controller 206-406 is in communication with an external user device. For example, the controller 206-406 may be in wireless communication with a smartphone, smart watch tablet or laptop of a user via the transceiver, although that is not essential and the controller 206-406 may alternatively be in wired communication with the external user device. The controller 206-406 may be configured to receive and transmit instructions and / or data to and from the external user device. That may allow the user to provide instructions to the controller 206-406 via the external user device to control some aspects of operation of the treadmill 100, for example initiating and deactivating operation of the treadmill 100 at the beginning and end of use, or selecting a virtual environment to be displayed on a display screen of the treadmill 100. That may also allow the user to receive or transmit data between the controller 206-406 and the external user device, for example relating to performance metrics relating to use of the treadmill 100 by the user (such as speed and / or inclination of the running surface 102 and / or distance travelled by the user on the running surface 102 during a session which may then be visualised or stored on the external user device, a target distance and time input by the user via the external user device to set a speed and duration of operation of the running surface 102 etc.). A software application or “app” may be provided on the external user device to provide a user interface via which a user can provide input to cause instructions and / or data to be exchanged between the controller 206-406 and the external user device. Alternatively, the external user 15 device may be a dedicated hardware device configured to interact or communicate exclusively with the controller 206-406, rather than a general purpose user device (such as a smartphone, tablet or laptop). Figure 5 shows an embodiment of a method 500 for controlling a treadmill such as the treadmill 100 shown in and described with respect to Figures 1, 3A and 3B. The method 500 may be performed using the control system 200, 300, 400 described above with respect to Figures 2, 3 and 4. Step 502 of the method 500 comprises obtaining image data of a user on the treadmill 100. Step 504 of the method 500 comprises providing the obtained image data to a machine learning algorithm trained to determine falling of the user from the image data. The obtained image data may be provided to a controller 206-406 on which the machine learning algorithm is provided. Step 506 of the method 500 comprises stopping movement of a running surface 102 of the treadmill 100 if falling of the user is determined by the machine learning algorithm. A control signal generated by the controller 206-406 on determining falling of the user based on the obtained image data may be used to stop movement of the running surface 102. It will be appreciated the method 500 may comprise one or more additional steps corresponding to one or more functions described with respect to the control systems 200-400 described above. From reading the present disclosure, other variations and modifications will be apparent to the skilled person. Such variations and modifications may involve equivalent and other features which are already known in the art of treadmill control systems, in particular control systems for treadmill safety, and which may be used instead of, or in addition to, features already described herein. Although the appended claims are directed to particular combinations of features, it should be understood that the scope of the disclosure of the present invention also includes any novel feature or any novel combination of features disclosed herein either explicitly or implicitly or any generalisation thereof, whether or not it relates to the same invention as presently claimed in any claim and whether or not it mitigates any or all of the same technical problems as does the present invention. Features which are described in the context of separate embodiments may also be provided in combination in a single embodiment. Conversely, various features which are, for brevity, described in the context of a single embodiment, may also be provided separately or in any suitable sub-combination. The applicant hereby gives notice that new claims may be formulated to such features and / or combinations of such features during the prosecution of the present application or of any further application derived therefrom. For the sake of completeness, it is also stated that the term “comprising” does not exclude other 5 elements or steps, the term “a” or “an” does not exclude a plurality, and any reference signs in the claims shall not be construed as limiting the scope of the claims.

Claims

29 04 261. A treadmill control system for use with a treadmill comprising:a camera configured to obtain image data of a user on a treadmill;5 means for detecting a position of the user on the treadmill; anda controller configured to receive the image data from the camera, and to receive data indicative of a position of the user on the treadmill;wherein the controller comprises a machine learning algorithm trained to determine falling of the user from the image data;10 wherein the controller is configured to determine if the position of the user on thetreadmill is within a pre-determined distance of a selected location on the treadmill; and wherein the controller is configured to generate a control signal configured to stop movement of a running surface of the treadmill on determining both:falling of the user; and15 a position of the user on the treadmill being within a pre-determineddistance of the selected location on the treadmill.

2. The treadmill control system of claim 1, wherein the selected location on the treadmill is at or adjacent a rearward end of the running surface of the treadmill.

203. The treadmill control system of any preceding claim, wherein:the system comprises a sensor configured to detect performance of one or more gestures by the user on the treadmill; andthe controller is further configured to:25 receive data from the sensor;determine performance of one or more pre-determined gestures by the user on the treadmill from the data; andgenerate, on determining performance of a pre-determined gesture by the user on the treadmill, a control signal to control operation of the treadmill.

304. The treadmill control system of claim 3, wherein the sensor configured to detect performance of one or more gestures comprises the camera configured to obtain image data of the user on the treadmill.35 5. The treadmill control system of claim 3 or of claim 4, wherein the controller comprisesa machine learning algorithm trained to determine performance of the one or more predetermined gestures by the user on the treadmill from the sensor data.29 04 266. The treadmill control system of claim 5, wherein the controller comprises a machine learning algorithm trained to:determine falling of the user from the image data; anddetermine performance of the one or more pre-determined gestures by the user on the treadmill from the sensor data.

7. The treadmill control system of any of claims 3 to 6, wherein the processor is configured to generate a control signal to control one or more of:a speed of the running surface of the treadmill; andan inclination of the running surface of the treadmill.

8. The treadmill control system of any preceding claim, wherein the means for detecting a position of the user on the treadmill comprises one or more of:a range sensor configured to provide data indicative of a distance between the user and the selected location on the treadmill; anda boundary sensor configured to provide data indicative of whether the user has reached the selected location on the treadmill.

9. The treadmill control system of any preceding claim, wherein the system does not comprise a manual user interface for receiving user instructions.

10. A treadmill comprising the treadmill control system of any preceding claim.

11. The treadmill of claim 10, comprising a display screen configured to display a virtual environment through which the user on the treadmill can navigate.

12. The treadmill of claim 10 or of claim 11, wherein a structure of the treadmill at or adjacent a forward end of the running surface consists essentially of a display screen.

Images