METHOD FOR LOCATING A PRE-PROGRAMMED AUTONOMOUS ROBOTIC VEHICLE OPERATING INSIDE AN O&G WELL, AND SYSTEM THAT EXECUTES SAID METHOD

Description

1 / 21 METHOD FOR LOCATING A PRE-PROGRAMMED AUTONOMOUS ROBOTIC VEHICLE OPERATING INSIDE AN O&G WELL, AND SYSTEM THAT EXECUTES SAID METHOD FIELD OF APPLICATION

[0001] This patent report concerns a method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well, and a system that executes said method, to enable analysis and interpretation of data obtained in real time by incorporated sensors, relating to relevant parameters, for autonomous decision making. SUMMARY OF THE INVENTION

[0002] Accurately identifying the positioning of autonomous tools within oil and petroleum product production or transportation pipelines has been a challenge in monitoring the integrity of these infrastructures.

[0003] Traditionally, the location of mobile devices in closed systems is obtained through odometers, which have limitations due to mechanical failures, such as slippage or locking, that can generate both punctual and cumulative errors in estimating the position of the equipment.

[0004] The method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well, and the system that executes said method, now proposed, aims to overcome these limitations by adopting a more precise approach based on the recognition of structural elements present in the pipeline, such as sleeves, connections, valves, among others. BACKGROUND OF THE TECHNIQUE Petition 870250000630, dated 05 / 01 / 2025, page 12 / 42 2 / 21

[0005] Currently, the depth / location data of a pre-programmed autonomous robotic vehicle operating inside an O&G well is provided by an odometer, which is susceptible to errors due to slippage / slipping during the movement of the equipment.

[0006] Therefore, this data should eventually be corrected with reference to data from a CCL (Casing Collar Locator) type sensor, which is a widely used methodology in oil well operations, measuring variations in the electromagnetic field along the well's tubing string. The reading, as it passes through the tubing sleeves, provides a response signal of the electromagnetic field variation that, at a given speed, generates a characteristic pulse (signature), which is analyzed and identified by the technician / engineer on the surface as being a tubing coupling sleeve.

[0007] Because the pipes installed in the well are measured beforehand, the distance of each sleeve from the surface is known, and this serves as a reliable reference for estimating the depth / location of the pre-programmed autonomous robotic vehicle, operating inside an O&G well, in relation to the bottom of the well.

[0008] Patent document CN115628047 discloses a CCL magnetic positioning tool, suitable for ultra-large wellbore holes, which can function normally in 13 to 24 inch diameter tubing, generating obvious coupling signals to obtain the depth position of an instrument, and which includes a CCL assembly, a flat spring assembly and a shaft, one end of the shaft being provided Petition 870250000630, dated 05 / 01 / 2025, page 13 / 42 3 / 21 with a socket assembly, and the other with a plug assembly, the socket assembly and the plug assembly being connected and communicating through the shaft, and a connecting rod assembly provided on the outer periphery; the shaft and the connecting rod assembly, including a fixed portion on the sliding part of the shaft, where several flat spring assemblies are installed, which are compressed against the inner wall of the well casing, to keep them in contact at all times, parallel to the metal shaft in the connecting rod assembly, the flat spring assembly actuating the sliding part to move relative to the fixed part, so that the CCL assembly maintains a certain distance from the flat spring assembly for measuring the bottomhole depth, the CCL assembly being electrically connected to the socket assembly for communication.

[0009] The specific problem of the state of the art lies in how to make an autonomous robotic apparatus, lowered into an O&G well, without communication with the surface, capable of reading and interpreting, in real time, the CCL data, correlating the depth in a completely autonomous manner, and thus making command decisions based on its precise depth.

[0010] As previously stated, currently, the data obtained in real time through the CCL are analyzed and interpreted in the operating cabin located on the surface by an engineer / technician, who identifies the sleeves and correlates the depth, comparing it with the previous schematic (tally) of the well string, with the aid of software.

[0011] Although the CCL is a standard logging tool adopted in the O&G industry for depth correlation in wells, all conveyance methods Petition 870250000630, dated 05 / 01 / 2025, page 14 / 42 4 / 21 (wireline, slickline and coiled tubing / flexitube) have odometry based on mechanical systems installed in the surface apparatus, which present inherent errors.

[0012] Therefore, it becomes desirable to propose a method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well, and a system that executes said method, to enable analysis and interpretation of data, obtained in real time, by incorporated sensors, referring to relevant parameters, for autonomous decision making. FUNDAMENTALS OF THE INVENTION

[0013] The method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well, and the system that executes said method, proposes to solve the drawbacks of current techniques by employing hardware containing a set of electronic circuits, which uses embedded firmware based on a deterministic finite state automaton, that treats and processes, in real time, the data from the CCL, through an initial treatment (statistical filter), to segregate noise inherent in the electromagnetic field reading, after which, the filtered data are analyzed to identify patterns / signatures with typical characteristics and aspects representative of production pipe sleeves, in the ascending / descending reading directions, to consider the direction of movement, and correct quantification and accounting of the sleeves.

[0014] Signatures are processed and used for correlation and depth correction with reference to the column schematic (tally), pre-loaded in the logic of Petition 870250000630, dated 05 / 01 / 2025, page 15 / 42 5 / 21 firmware, which ultimately corrects the depth in case of a systemic odometer error.

[0015] The proposed method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well allows the identification of various magnetic signature patterns from CCL sensor data collected along the tool's path through the production string.

[0016] The method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well adapts to the nuances, as well as discrepancies, found in the various patterns of magnetic signatures identified.

[0017] The method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well is capable of analyzing different chronological patterns of magnetic signatures linked to a specific well structure (tubing couplings, for example), distinguishing, in real time, the direction traveled by the tool.

[0018] The method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well allows loading previously acquired signature reference patterns in accordance with the structural diversity expectations contained in a particular well, and also allows the addition of new signature patterns to meet potential incipient demands.

[0019] The method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well also identifies signatures without the need for prior data from the wells under analysis. Petition 870250000630, dated 05 / 01 / 2025, page 16 / 42 6 / 21

[0020] The method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well was designed for both surface-controlled tools and fully autonomous robotic devices.

[0021] The method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well, and the system executing said method, a device constitutes a system dedicated to the automatic location and depth correlation of a logging tool or autonomous robotic unit operating inside a well.

[0022] The method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well allows for concurrent odometric correction and synchronization of other embedded systems and / or sensors from third-party devices.

[0023] The method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well operates in real time, interpreting signatures locally and allowing any autonomous robotic device to know and correlate its depth / location relative to the well.

[0024] The method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well allows computing the distance traveled and accumulated, for the purpose of calculating the remaining range for fully autonomous robotic systems or logging tools.

[0025] The method for locating a pre-programmed autonomous robotic vehicle operating inside a well Petition 870250000630, dated 05 / 01 / 2025, page 17 / 42 7 / 21 O&G allows computing the current depth of the tool, in order to ensure the depth / location of a specific pre-determined point for intervention operations by fully autonomous robotic systems or profiling tools.

[0026] The method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well includes a communication interface that allows raw and / or interpreted data, collected locally, to be sent to remote devices, in the case of application with a well logging tool.

[0027] The identification of magnetic signature patterns of structures contained in the well is designed with scalability, allowing the addition of recognition of new signature patterns.

[0028] The method for locating a pre-programmed autonomous robotic vehicle, operating inside an O&G well, can store the historical record of the raw data collected during its excursion in the well.

[0029] The method for locating a pre-programmed autonomous robotic vehicle, operating inside an O&G well, is implemented as a solution of multiple deterministic finite state machines, coexisting in an embedded application and associated with particular signatures.

[0030] The method for locating a pre-programmed autonomous robotic vehicle, operating inside an O&G well, involves the use of artificial intelligence, when the robotic vehicle is pre-programmed from the surface.

[0031] The method for locating a pre-programmed autonomous robotic vehicle operating inside a well Petition 870250000630, dated 05 / 01 / 2025, page 18 / 42 8 / 21 O&G is a mix of statistical and deterministic algorithms in the embedded solution.

[0032] The method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well includes a sensor fusion strategy to increase the accuracy of its interpretations.

[0033] The method for locating a pre-programmed autonomous robotic vehicle, operating inside an O&G well, is able to exclude non-statistical components contained in the signal. BRIEF DESCRIPTION OF THE FIGURES

[0034] For a better understanding of the method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well, reference is made to the attached figure, in order to characterize its functionality; however, this figure is merely illustrative and may vary, provided it does not deviate from its functional principle, and where: Figure 1 presents the block diagram of the method for locating a pre-programmed autonomous robotic vehicle operating inside an oil and gas well, of the present invention; Figure 2 shows a typical magnetic signature, obtained by a CCL sensor in transit through a segment of the well; Figure 3 presents groups of possible signature sequences; Figure 4 shows the deterministic state transition diagram; Petition 870250000630, dated 05 / 01 / 2025, page 19 / 42 9 / 21 Figure 5 presents an improved version of the deterministic state transition diagram of the automaton in Figure 3; Figure 6 shows the flow of states through the evolution of an advance movement handled by the automaton in Figure 4; Figures 7 to 9 show state flows through the evolution of a recoil movement handled by the automaton in Figure 4. PREFERRED DESCRIPTION OF THE INVENTION

[0035] In accordance with what is illustrated in the block diagram of Figure 9, the method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well begins from the actual position (depth measured in the production string / pipes) of the device in the well, and as the device moves through the well, data is acquired by the sensors: Glove locator 1; Glove locator 02 (optional); odometer; . pipe end locator / pipe inside diameter gauge (optional).

[0036] The data from glove locator sensors 1 and 2 undergo statistical processing / filtering, which serves to filter and segregate the inherent noise from reading the electromagnetic signal collected during movement along the tubes.

[0037] The filtered data then undergoes processing and interpretation by a deterministic finite-state automaton to identify characteristic patterns / signatures. Petition 870250000630, dated 05 / 01 / 2025, page 20 / 42 10 / 21 with typical and representative aspects of production pipe sleeves, in both reading directions (ascending and descending) for determining the direction of movement and correct quantification and accounting of the sleeves.

[0038] Signatures are processed in the correlation control system for depth correction with reference to the column diagram / map preloaded in the firmware logic.

[0039] In case of a systemic error in the depth value originally provided by the odometer, it is corrected based on the correlated value, which becomes the corrected real position in the well.

[0040] The central idea of ​​the method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well is to identify the equipment's location by detecting specific magnetic characteristics of the pipeline components. This is done using a CCL-type magnetic sensor, which captures variations in the magnetic field around the equipment, generated by the pipeline and well elements.

[0041] The CCL sensor can identify distinct magnetic patterns, which are correlated with a pre-loaded map in the system's memory, and which contains information about the precise depths measured sequentially of the piping elements, such as connections and valves, for comparison of the signals captured by the sensors with the map data, to determine their relative position.

[0042] Thus, the central idea is based on the use of specific automata correlated to the elements, for the identification of the signature of each element. Petition 870250000630, dated 05 / 01 / 2025, page 21 / 42 11 / 21

[0043] For the automaton to effectively identify patterns or signatures of elements in the piping, it is essential that the CCL sensor signal be properly pre-processed and filtered to enhance important signal characteristics such as peaks, valleys, widths, and amplitudes, making it more distinct and less susceptible to noise or interference. Approaches such as low-pass, band-pass, median filter, Savitzky-Golay filter, Kalman filter, particle filter, wavelet transform, adaptive filters such as LMS and RLS, FFT, and others can be used.

[0044] The method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well, as proposed herein, was conceived considering the need for simplicity in embedded solutions applicable in the well environment, due to processing constraints and high MTTF requirements typical of aggressive environments.

[0045] Its implementation results in low latency, fast response, extremely low energy consumption, and the expectation of high operating time.

[0046] The signal from a CCL sensor is subjected to two processing stages, the first being aimed at suppressing noise contained in the signatures and statistical calculations, the results of which produce an output signal suitable for the deterministic processing of the second block of the solution, formed by deterministic finite automata, both synchronous and asynchronous.

[0047] The proposed solution operates predominantly in asynchronous mode, responding to events contained in the signal, and may Petition 870250000630, dated 05 / 01 / 2025, page 22 / 42 12 / 21 can be implemented only by software, hardware, or a combination of both.

[0048] The methodology employed to ensure the scalability and coexistence of multiple automata is based on self-exclusion, and does not require an arbitration agent to guarantee its robustness.

[0049] Although it is assumed, from this point on, that the practical applications of the method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well are restricted solely to the characteristic signatures of the gloves, such an approach should not limit the scope of the present invention to this specific type of application.

[0050] Figure 2 shows a typical magnetic signature, obtained by a CCL sensor in transit, through a segment of the well, making it possible to identify a total of five signatures, corresponding to five sleeves arranged along the path.

[0051] To facilitate understanding, the peak values, identified around the average value, have been numbered, and it can be seen below the graph that the peaks have been grouped according to the signature pattern to which they belong.

[0052] The signatures were labeled (Signature pattern) in order to associate the peaks with their respective clustering, which represents the well structure observed in the graph.

[0053] The use of symbols > and < indicates that the samples contained in the signature are located, respectively, above or below the range of values ​​delimited by the upper (TH+) and lower (TH-) thresholds. Petition 870250000630, dated 05 / 01 / 2025, page 23 / 42 13 / 21

[0054] Sample data outside this range are relevant for processing by the automaton; glove signature patterns may be <, >, <, indicating the device's advance through the glove, or >, <, > in the case of retraction.

[0055] By chronologically arranging the symbols > and <, associated with the peaks / pulses contained throughout the graph, it is possible to extract and identify four groups of possible sequences, each containing groupings of three peaks, which have been given mnemonics representing the type of movement in which they are obtained; they are: Forward motion (<, <, >); . Forward inversion (<, >, <); Reverse motion (>, >, <); . Reverse inversion (>, <, >);

[0056] It is important to understand that there is a rule in the succession between distinct sequences as shown in figure 3.

[0057] For reasons of symmetry, the point of passage through the sleeve is computed, for any direction in the displacement, if it occurs at the central position of the element under analysis, in the example given, the sleeve; however, this method does not exclude the hysteresis operation, where the increment or decrement of the number of structural elements identified in the signature occurs at the transition between the edges of its structure and the production column.

[0058] Considering the computing pattern adopted in figures 2 and 3, the peaks in the signatures corresponding to the central point are identified in gray at positions 1, 4, 7, 10 and 13 in figure 3.

[0059] To determine the direction in which the CCL device travels through the sleeve, it is necessary to know that Petition 870250000630, dated 05 / 01 / 2025, page 24 / 42 14 / 21 polarity of the last pulse recorded when passing through the previous glove, and, for this reason, at the peak in Figure 2 it is not possible to determine the advance through the first glove without additional resources.

[0060] Based on the information shared so far, Figure 4 presents the state transition diagram, where, in addition to the relationships between states (inputs x transitions), names have been assigned to them, and, as is known, in a finite automaton each state is a numerical representation of the record of relevant past events, and, based on the current point and the present inputs, the system advances through a number of possible states pre-established in the design. The legend of Figure 4 links the mnemonic assigned to the nature of the movement conferred to the sensor, with the flow of states assumed during its recognition.

[0061] For obvious reasons, the automaton in Figure 4 is a didactic implementation and does not consider the discretization resulting from the data provided by the CCL.

[0062] In Figure 2, the signature shown is the result of interpolating data from successive discrete samples in time and amplitude, which already demonstrates the need for correction in the diagram of the first automaton, since it treats a pulse as a single event, and not as a packet of interpolated samples; thus, successive samples within the same pulse, classified as exceeding in magnitude one of the noise discrimination thresholds (TH+ and TH-), are individually computed as isolated peaks and result in undesirable transitions between states, because they behave like a pulse train. Petition 870250000630, dated 05 / 01 / 2025, page 25 / 42 15 / 21

[0063] Figure 5 shows an improved version of the same automaton as in Figure 4, highlighting the need to include additional deterministic states.

[0064] The enhanced automaton in Figure 5 can be started in four possible states, considering the direction of movement to which the device will be subjected and also its arrangement, inside or outside the column, at the moment the operation begins.

[0065] In figure 5, the flow of states through the evolution of an advance movement handled by the automaton in figure 5 is presented superimposed on the signature corresponding to this type of movement.

[0066] State B1 is reached when, with the device in initialization state A3, a sequence of one or more samples with a value lower than the TH- threshold is recorded, and it is preserved as long as this condition is maintained.

[0067] State B2 represents the completion of reception of the first of three pulses that make up an advance sequence and succeeds state B1 at the moment when the magnitude assumed by the sample exceeds the lower noise discrimination threshold, and the automaton will remain in B2 as long as successive samples from the signature have an amplitude between the established thresholds TH+ and TH-.

[0068] The automaton progresses to state C1 when the sample value from the signature falls below the TH- threshold, representing the computation of the second (negative) pulse of a successful sequence characteristic for forward movement, and the assumption of this state has an evolutionary representation of the history of necessary events. Petition 870250000630, dated 05 / 01 / 2025, page 26 / 42 16 / 21 to the positive validation of what is interpreted as an advance shift, being in C1 and as long as samples originating from CCL have a value lower than TH-, the automaton will remain indefinitely in the same state, since such samples are components of this same second pulse.

[0069] State C2 represents the recording of the termination of the second of three expected pulses.

[0070] The A2 state is established when the current sample in the signature reaches a value higher than the TH+ threshold, which is interpreted as a positive pulse closing a characteristic magnetic signature of advancement; subsequent input samples that remain above the TH+ level ensure the maintenance of this same state by understanding that they are component elements of this same pulse.

[0071] In A2, the advance is accounted for, which translates to the updating of the output produced by the automaton.

[0072] State A2 is only succeeded by A3 when a sample at the input has an amplitude level in the range between TH+ and TH-, re-establishing the initial state.

[0073] Similar reasoning can be applied to the movements represented by figures 7, 8 and 9.

[0074] As a second example, consider the flow of transitions between states associated with the pullback movement initiated by state E3, which is achieved by two possible paths; the first occurring when a successful sequence of pulses, characteristic of a previous pullback movement, promotes a flow of states culminating in the termination of the cycle by E3, and the second when the sequence called forward reversal occurs, which appears at the moment when the Petition 870250000630, dated 05 / 01 / 2025, page 27 / 42 17 / 21 The direction of forward movement is reversed, identified through the recording of an appropriate sequence of states, ending in the stabilization of E3. In a backward movement, the machine states progress in a way that recalls the relevant history of events, where state F1 represents the recording of a pulse with positive polarity, considering the average value as a reference, and remains in it as long as successive samples from the CCL sensor remain with an amplitude above the upper noise discrimination threshold (TH+), given that such samples would be component elements of this same pulse.

[0075] State F2 represents the termination of the pulse recorded by state F1, and F2 is reached when the sample obtained from the CCL returns to the amplitude range of the signal that represents the noise; this range is dynamic and updated with each interaction of the statistical module, and the system remains in this state until one of the established thresholds is reached.

[0076] The H1 state is the memory record of the successful reception of the second of the three pulses that make up the magnetic signature at recoil, and the maintenance of the automaton in this state occurs concomitantly with the reception of samples whose amplitudes exceed the dynamic threshold TH+.

[0077] When the pulse ends, the sample amplitude reaches a value less than TH+, being recorded with the transition of the machine from state H1 to state H2, which can be interpreted as the completion of the recording of the first two pulses contained in the signature, associated with the recoil movement of the sensor. Petition 870250000630, dated 05 / 01 / 2025, page 28 / 42 18 / 21

[0078] With the automaton having assumed state H2, and with the samples from the CCL having amplitudes lower than the dynamic threshold TH-, the system is driven to state E1, whose function is to compute the reception of a negative pulse, considering the average value as a reference, remaining in it as long as the successive samples at the input maintain amplitudes lower than the threshold TH-, which represent part of the same pulse, and in E1 the system computes the decrement of the accumulated glove value by -1.

[0079] The occurrence of a sample that reaches a value greater than TH- is understood as the end of the third pulse of the signature and implies a new register transition, representing the restart of the cycle, the E3 state.

[0080] To increase accuracy and reduce errors caused by variations in the magnetic field or external interference, an additional variant of the system is proposed for executing the method for locating a pre-programmed autonomous robotic vehicle operating inside an O&G well, with an alternative topology to using only one sensor, which uses two or more CCL sensors positioned at a defined and rigid distance from each other.

[0081] The idea is that, when adopted, the combination of multiple sensors allows for a more robust and reliable reading of magnetic signals, as well as facilitating data processing to correct any disturbances or noise in the signal; in addition, the use of more than one sensor enables the use of signal processing techniques, such as the averaging of received signals, or the application of statistical filters to smooth the readings, thus increasing the accuracy of the positioning. Petition 870250000630, dated 05 / 01 / 2025, p. 29 / 42 19 / 21

[0082] Alternatively, velocity calculation can be used to determine the position of the equipment along the pipeline, working as follows: Sensor readings: the first sensor, positioned at the front of the equipment, detects a magnetic event, which could be a change in the magnetic field caused by an element in the piping, such as a valve or connection, and the system records the time of detection of this event; Detection by the second sensor: the second sensor, positioned at a fixed distance from the first, detects the same or a similar magnetic event as the equipment moves along the piping; Speed ​​calculation: the system calculates the time between the detection of the two signals, using the fixed distance between the sensors and the recorded time between readings, and, from this data, the speed of the equipment can be determined; Position calculation: with the calculated velocity, it is possible to estimate the position of the equipment from an initial point; the system integrates the velocity over time to estimate the distance traveled and determine the relative position of the device within the pipeline.

[0083] The system that executes the method for locating a pre-programmed autonomous robotic vehicle, operating inside an O&G well, offers the following technical advantages: When using multiple CCL sensors, instead of a single sensor, it reduces the likelihood of errors due to sensor failures or external interference, resulting in a more accurate positioning system; Petition 870250000630, dated 05 / 01 / 2025, pages 30 / 42 20 / 21. Eliminating reliance on odometers, which are susceptible to mechanical failures, offers a more robust alternative for determining the equipment's position; When using multiple sensors, advanced signal processing is possible, which helps filter out noise or variations in the magnetic field that can affect the accuracy of a single sensor's reading. . allows the equipment to move autonomously and intelligently, adapting to the piping and using contextual information about the environment to make directional decisions and locate itself precisely; The correlation of magnetic signals with a digital map of the pipeline allows the system not only to determine the position of the equipment, but also to relate that position to specific pipeline integrity events, such as faults or defects, optimizing data collection for analysis.

[0084] In order to correct cumulative errors caused by uncertainty in velocity measurement, the system can adopt several complementary approaches that increase the robustness of the position calculation and minimize the impact of small variations over time, such as: Periodic recalibration using known standards whenever the equipment passes through a known reference point (such as a valve or pipe connection with well-defined magnetic characteristics), and it can also recalibrate its position by comparing the current reading with the pre-loaded map; Data filtering (Kalman filter) combines imprecise measurements from multiple sensors to estimate position. Petition 870250000630, dated 05 / 01 / 2025, page 31 / 42 21 / 21 based on a mathematical model of the system's dynamics, which can be adjusted to correct deviations based on velocity uncertainties and detect possible flaws or inconsistencies in the data; . Data fusion with additional sensors, beyond the CCL sensors, promoting the integration of other types of sensors, such as gyroscopes, accelerometers, or pressure sensors, to improve position estimation and correct speed errors, providing additional data on the acceleration or orientation of the equipment, to reduce cumulative error; Continuous displacement verification, allowing for the implementation of a continuous self-checking strategy for each section of the pipeline, where CCL sensor signals are analyzed at regular intervals to detect inconsistencies or unexpected patterns that may indicate accumulated errors, so that, by identifying discrepancies between the expected and the actual values, the system can correct its position estimate and adjust the calculated speed. Petition 870250000630, dated 05 / 01 / 2025, pages 32 / 42

Claims

1 / 6 CLAIMS 1. METHOD FOR LOCATING A PRE-PROGRAMMED ROBOTIC AUTONOMOUS VEHICLE OPERATING INSIDE AN O&G WELL, characterized by starting from the actual position of the device in the well, relative to the depth measured in the production string / pipes, and, as the device moves through the well, data is acquired by the following sensors: . sleeve locator 1; . sleeve locator 2 (optional); . odometer; . end-of-pipe locator / pipe internal diameter gauge (optional).

2. METHOD, according to claim 1, characterized in that the data from the glove locator sensors 1 and 2 undergo statistical treatment / filtering to filter and segregate the inherent noise from the electromagnetic signal reading collected during movement along the tubes.

3. METHOD, according to claim 1, characterized by the fact that the filtered data undergo processing and interpretation by a deterministic finite state automaton to identify characteristic patterns / signatures with typical and representative aspects of production pipe sleeves, in the ascending and descending reading directions, to determine the direction of movement and correct quantification and accounting of the sleeves.

4. METHOD, according to claim 1, characterized by the fact that the signatures are processed in the correlation control system for depth correction with reference to the column diagram / map preloaded in the firmware logic.

5. METHOD, according to claim 1, characterized in that, in case of a systemic error in the depth value originally provided by the odometer, this is corrected based on the correlated value, which then becomes the corrected actual position in the well.

6. METHOD, according to claim 1, characterized in that the detection of specific magnetic characteristics of the piping components is performed by a CCL (Casing Collar Locator) type magnetic sensor.

7. METHOD, according to claim 1, characterized in that the distinct magnetic patterns captured by the CCL sensor are correlated with a map pre-loaded in the system's memory, which contains information about the precise depths measured sequentially of the pipe elements.

8. METHOD, according to claim 7, characterized in that the CCL sensor can detect disturbances in the magnetic field related to points in the piping.

9. METHOD, according to claim 1, characterized by the fact that it uses specific correlated automata to identify the signature of each element.

10. METHOD, according to claim 1, characterized by the fact that the automaton identifies patterns or signatures of elements in the piping from the enhancement of important signal characteristics, such as peaks, valleys, widths and amplitudes, making it more distinct, and less susceptible to noise or interference.

11. METHOD, according to claim 10, characterized by the fact that it can utilize approaches such as, for example, low-pass, band-pass, median filter, Savitzky-Golay filter, Kalman filter, particle filter, wavelet transform, adaptive filters such as LMS and RLS, FFT, and others.

12. METHOD, according to claim 1, characterized by taking into account processing constraints and high MTTF requirements.

13. METHOD, according to claim 1, characterized by having low latency, fast response, extremely low power consumption, and an expected high operating time.

14. METHOD, according to claim 1, characterized in that the deterministic finite automata are both synchronous and asynchronous.

15. METHOD, according to claim 14, characterized in that the operation, predominantly in asynchronous mode, responding to events contained in the signal, can be implemented only by software, hardware or a combination thereof.

16. METHOD, according to claim 1, characterized in that the scalability and coexistence of multiple automata is based on self-exclusion, and does not require an arbitration agent to guarantee its robustness.

17. SYSTEM FOR EXECUTING THE METHOD FOR LOCATING A PRE-PROGRAMMED ROBOTIC AUTONOMOUS VEHICLE, Petition 870250000630, dated 05 / 01 / 2025, page 38 / 42 4 / 6 OPERATING INSIDE AN O&G WELL, for executing the method described in any of claims 1 to 16, characterized in that, to increase accuracy and reduce errors caused by variations in the magnetic field or external interference, it uses one or more CCL sensors positioned at a defined and rigid distance from each other.

18. SYSTEM, according to claim 17, characterized in that the use of more than one sensor enables the use of signal processing techniques, such as the averaging of received signals, or the application of statistical filters to smooth the readings, increasing the accuracy of the positioning.

19. SYSTEM, according to claim 17, characterized in that, alternatively, the speed calculation can be used to determine the position of the equipment along the pipeline, functioning as follows: . the first sensor, positioned in front of the equipment, detects a magnetic event which may be a change in the magnetic field caused by an element of the pipeline, such as a valve or connection, and the system records the time of detection of this event; . the second sensor, positioned at a fixed distance from the first, detects the same or a similar magnetic event as the equipment moves along the pipeline; . the system calculates the time between the detection of the two signals, using the fixed distance between the sensors and the time recorded between the readings, and, from these data, the speed of the equipment can be determined; Petition 870250000630, dated 05 / 01 / 2025, p. 39 / 42 5 / 6.With the calculated speed, it is possible to estimate the position of the equipment from an initial point; the system integrates the speed over time to estimate the distance traveled and determine the relative position of the device within the pipeline.

20. SYSTEM, according to claim 17, characterized in that, to correct cumulative errors caused by uncertainty in speed measurement, complementary approaches can be adopted that increase the robustness of the position calculation and minimize the impact of small variations over time, such as: . periodic recalibration using known standards whenever the equipment passes through a reference point with defined magnetic characteristics, to recalibrate its position, comparing the current reading with the preloaded map; . data filtering by combining imprecise measurements from multiple sensors and estimating the position based on a mathematical model of the system dynamics, which can be adjusted to correct deviations based on speed uncertainties and detect possible flaws or inconsistencies in the data; .Data fusion with additional sensors, beyond the CCL sensors, promoting the integration of other types of sensors, such as gyroscopes, accelerometers, or pressure sensors, to improve position estimation and correct speed errors, providing additional data on the acceleration or orientation of the equipment, to reduce cumulative error; Petition 870250000630, dated 05 / 01 / 2025, pp. 40 / 42 6 / 6. Continuous displacement verification, being able, for each section of the pipeline, to implement a continuous self-verification strategy, where the signals from the CCL sensors are analyzed at regular intervals to detect inconsistencies or unexpected patterns that may indicate accumulated errors, so that, upon identifying discrepancies between the expected and the actual, to correct its position estimate and adjust the calculated speed. Petition 870250000630, dated 05 / 01 / 2025, pp. 41 / 42.

Images