EP3060480A1 - Collaborative robot for visually inspecting an aircraft - Google Patents

Abstract

The invention relates to a device (100) for visually inspecting the external surfaces (90) of an aircraft, which comprises an inspection area for receiving an aircraft, at least one visual inspection robot (10), and a control center (50). A movable platform (11) of the robot supports a turret (12) having observing means (13). The robot comprises processing means (20) which guide the movable platform (11) and process the data received from the observing means (13). The processing means (22) of the robot are suitable: for autonomously controlling the robot (10) during the visual inspection of the external surfaces of an aircraft (9) parked in the inspection area; for interrupting a visual inspection in the event of a detection of an anomaly on the external surface of the aircraft; for transmitting visual inspection data to the control center; and for receiving instructions from the control center.

Description

 Collaborative Visual Inspection Robot of an Aircraft

The present invention belongs to the field of non-destructive testing of aircraft.

 More particularly, the invention relates to a collaborative robot for visually inspecting an aircraft on the ground as part of the verification or control operations and relates to an inspection method implementing such a robot. The invention finds particular application in the field of control operations before the flight of an aircraft.

In the field of aircraft, the visual inspection, before flight or during maintenance operation, is included in the periodic checks that ensure the operational safety of the aircraft.

 Indeed visual inspection can detect anomalies that can occur on visible parts of the aircraft that are often the most exposed to outside influences and in some cases can reveal deeper damage to the structure.

 In addition visual inspection does not implement any particular disassembly, at most the opening of doors or hatches which allows a relative speed of inspection. Historically and still largely today, the visual inspection is performed by a ground operator, who is for example a mechanic or a pilot of the aircraft.

 The inspection by a ground operator is carried out following a check list, but the operator can freely inspect other elements or areas of the aircraft than those provided in the checklist, particularly if Indices lead to an interest in a particular area, which improves the detection of possible anomalies.

On the other hand, an operator can be more or less attentive and be distracted observations that are not necessarily the most significant, at the risk of neglecting certain detail information that should be subject to in-depth review and analysis. To avoid the hazards of human intervention it was considered to carry out visual inspections automatically. For this purpose, systems using video cameras fixed with respect to the ground, for example in a hangar, or mobiles carried by robots moving on the ground or in flight, have been imagined. Patent application WO2012 / 047479 illustrates an example of an automated inspection device for an aircraft. The inspection of the aircraft is supported by a set of fixed or mobile cameras on the ground by means of mobile robots or mobile in flight by means of flying robots that are assigned to different parts of the aircraft. The display means communicate with a remote computer center which processes the received images to deduce the presence of anomaly and determines the maintenance operations to be performed.

 In such a system the processing of the images and their interpretations are performed automatically which allows the control operations to be performed without human intervention, the decision-making being entrusted to the automatic system. If the hazards of the human operator are thus avoided, it is then lost the benefit of the observation made by an operator who has great capacity for interpretation, especially in the presence of new and unregistered situations whose interpretation can be difficult. In addition, a device that automatically controls security issues poses potential certification problems because of the necessary demonstrations of high system reliability.

To avoid the difficulties of the known solutions, the device of the invention for the visual inspection of the external surfaces of an aircraft comprises an inspection area intended to receive an aircraft and comprises at least one visual inspection robot including a flat mobile form carries a turret with display means and comprises processing means providing a guidance of the mobile platform and a processing of information received from visualization means. In addition, the visual inspection device comprises a control center with a station for at least one control operator, and the processing means of the visual inspection robot are adapted to:

 driving the visual inspection robot autonomously during a visual inspection of the exterior surfaces of an aircraft parked on the inspection area;

 discontinue a visual inspection in progress if an anomaly is detected on the outer surface of an aircraft being inspected;

 transmit visual inspection information to the control center;

 receive instructions from the control center for the follow-up to be given to a visual inspection.

 There is thus obtained a device for visually inspecting an aircraft entirely controlled by a remote operator assisted by a collaborative robot performing the visual inspection tasks in the vicinity of the aircraft.

 Advantageously, the visual inspection robot comprises means for determining at any time during an inspection the position of the visual inspection robot and the orientation of the display means in an axis system linked to the aircraft.

 The robot is thus able to move autonomously with respect to the aircraft both for the management of its movements and for that of the areas of the aircraft to be inspected visually.

 In one embodiment, the robot processing means are adapted to determine the position of the robot and the orientation of the display means by an image processing of the aircraft to be inspected obtained by the display means.

 Such a means makes it possible to readjust the position of the robot with respect to the aircraft even if the aircraft does not have an exact position with respect to the theoretical position which it should have and to correct the deviations of this position without measurement or means external to the device.

In one embodiment, the visual inspection robot has absolute location means, such as a GPS receiver or laser rangefinders. pointing reference targets, and or means for integrating its movements, for example by odometry.

 The robot is thus able to move independently of any registration on the position of the aircraft to get close to the aircraft and after registration to determine its precise position during its movements even without establishing its position continuously by direct observation of the aircraft.

 In addition, the visual inspection device comprises all or some of the following features:

 The processing means comprise data storage means comprising, at least temporarily, characteristics, in particular geometric and graphic, of an aircraft to be inspected. The robot thus locally has the nominal characteristics of the aircraft he must inspect and vis-à-vis which he must visually identify any anomalies.

 The processing means comprise data storage means comprising anomaly characteristics, for example in an anomaly library. The robot, in addition to logic engines to identify anomalies not necessarily listed, is thus able to compare any observed visual element with respect to known anomalies.

 The processing means comprise image processing algorithms for detecting visible anomalies in the images produced by the viewing means in at least one of the wavelengths of the optical spectrum.

 The display means comprise means of illumination in a light of the visible range and or in the field of the infrared and or in the field of the ultraviolet.

 It is thus not only illuminated areas of the aircraft that can naturally be dimly lit but also highlighted anomalies sensitive to certain light spectra or whose visual contrast is increased at certain wavelengths of the light spectrum.

The display means and the processing means are configured to determine a three-dimensional shape of the exterior surfaces inspected of the aircraft. It is thus possible to identify which forms of the outer surface of the aircraft do not conform to the nominal shapes.

 In one embodiment of the device, the visual inspection device comprises means of non-destructive control of the structure of the inspected aircraft.

 The non-destructive inspection means are carried by the visual inspection robot or are worn, in whole or in part, by at least one control robot whose behavior is controlled by the visual inspection robot.

 It is thus possible to confirm or deny that the visual anomaly is the index of a deeper structural anomaly and to measure non-visible damage.

 The display means are orientable in location and in azimuth with respect to a repository of the platform of the visual inspection robot. It is thus possible to quickly scan all the outer surfaces of the aircraft visible by the robot during its movement.

 In one embodiment, the processing means are configured to determine the position of a fault detected on an aircraft with respect to elements of the internal structure of said aircraft, which are not visible from outside the aircraft. The consequences of the defect are thus better evaluated and the position of the defect is localized for the maintenance operators in relation to identifiable structural elements of structure.

 For example, the inspection robot is a robot rolling on a floor of the inspection area or moving by levitation in a volume whose footprint substantially corresponds to the inspection area.

 In one embodiment of the device, a plurality of inspection robots configured to perform a visual inspection of the same aircraft is implemented. This results in a faster and, if necessary, more complete visual inspection if specialized robots to inspect certain areas are implemented.

The invention is also directed to a method of visually inspecting a aircraft in which images of an outer surface of the inspected aircraft are transmitted to processing means of a visual inspection robot, wherein the processing means analyzes the images to identify the presence of any visible anomalies , and wherein when a visible anomaly is detected, the detected anomaly data is transmitted to a control center and the visual inspection is interrupted, at least when the processing means identifies an abnormality belonging to a category of anomalies considered critical.

 Advantageously, when the inspection is interrupted because of the detection of an anomaly, instructions are transmitted to the visual inspection robot by the control center for the continuation of the visual inspection, said instructions determining how the robot must continue. inspection.

 In one form of implementation, the amplitude of a visible anomaly is calculated by the processing means from optical deformation measurement means and or by a colorimetric analysis in the visible range, and or infrared and / or ultraviolet of the light spectrum.

 In one embodiment, when an abnormality is visually detected, an area affected by the visible defect is subjected to non-destructive testing by the visual inspection robot or a non-destructive control robot controlled by the robot. visual inspection.

The invention is described with reference to the figures, which schematically shows a nonlimiting example embodiment of the invention:

 Figure 1: an example of a basic arrangement of the main components of a visual inspection device of an aircraft implementing a collaborative robot and a remote control center;

 Figure 2: an example of a flowchart of the main steps of the visual inspection process implementing a collaborative robot and a remote control center.

A collaborative inspection robot of an aircraft 90 such as the robot 10 illustrated in FIG. 1 comprises a mobile platform 11 provided with a turret 12 carrying viewing means 13.

The illustrated mobile platform 11 is mounted on four wheels ensuring the stability of the platform and its displacement on the ground by the motorization of at least one of the wheels.

Other arrangements to ensure mobility and stability of the platform ^¬ form, for example the implementation of more or less wheels, for example, three wheels or six wheels, or for example caterpillars, plain solutions are possible . The choice of these means is advantageously made according to the nature of a soil on which the robot 10 will have to move during the inspection operations around the aircraft 90.

 The turret 12 is secured in a lower portion of said turret of the platform 11, for example on an upper face of said platform, and the display means 13, in the illustrated example a digital camera, are integral with the turret 12 in an upper part of said turret.

All the mechanical links between the platform 11 and the turret 12, and between the turret 12 and the display means 13 are arranged so that a main direction of observation 131 of the display means 13 is steerable. site and in azimuth in all possible directions to aim points above the ground and advantageously also on the ground on which the robot 10 moves, if necessary taking into account the platform's displacement capabilities him providing the possibility of changing the azimuth of a line of faith 111 of said platform.

 For example, the turret 12 is made to modify the azimuth of the main direction of observation 131 with a limited angular amplitude, for example 180 °, the other directions being then reached by movements on the ground of the platform.

 Such an arrangement makes it possible in particular to observe extended areas in azimuth, including without movement of the platform.

The orientation capabilities in site of the main direction of observation 131 is advantageously at least 90 ° between a target substantially horizontal and a substantially vertical aim upward.

 Where appropriate, the substantially horizontal aiming allows a downward aiming to allow the observation from above of areas that may be located under the viewing means 13, placed for example in an elevated position by means of the turret 12. Such areas are for example parts of the ground under the aircraft likely to show signs of leakage of a fluid from the aircraft. Such zones are, for example, treads of tires likely to bear signs of abnormal wear or of parts of structure that can be observed from above because of the raised position of the display means.

In one embodiment, the turret 12 includes elevation means 121 of the display means for modifying the height above the ground of said display means.

Such elevation means 121 consist for example of an assembly of articulated arms, two or more arms, which unfold or fold by actuators, not shown, so that the height above the plate ^¬ form 11, and therefore above the ground, the upper end of the turret 12 carrying the display means 13 can be modified.

The display means 13, in an elementary form, consist of a camera providing images in the field of visible light and intended to be analyzed. The images can be transmitted in analogue form to be processed if necessary after being converted into digital form or transmitted in digital form.

 Advantageously, the display means 13 comprise illumination means 132, for example integral with the turret 12, oriented to illuminate areas to be inspected located in a field of view of said viewing means.

The illumination means 132 may be offset from the main viewing direction 131 of the display means 13 to create more or less grazing lights for viewing certain defects such as superficial depressions.

 The illumination means 132 may also be arranged to produce illuminations in particular areas of the visible light spectrum, of red, green or blue color, for example, or non-visible, corresponding to particular spectral bands of the infrared or ultraviolet spectra, and are advantageously switchable between the different visible or non-visible spectral bands.

 The spectra of the lights of the illumination means 132 are set according to the spectral sensitivities of the cameras used for the display means 13 and according to the expected behaviors of the zones inspected under the different lighting, in particular when contrast effects or fluorescence are potentially expected to visualize and / or diagnose an abnormality.

 In embodiments, the display means 13 comprise particular equipment, not shown in the drawings, such as for example stereoscopic cameras and or other devices for giving a three-dimensional perception, for example by stereo-correlation techniques. , shearography, laser telemetry ... The perception in three dimensions makes it possible to check the conformity and to detect the deformations of both curved surfaces and plane surfaces.

In a preferred embodiment, the mobile platform 11 is self-powered.

 The propulsion can be carried out by thermal engines ensuring the driving of the wheels for the displacement of the mobile platform, for example by a hydrostatic transmission, and to produce a hydraulic and / or electrical power necessary for actuators useful for the movements of the turret 12 and viewing means 13.

Propulsion can also be achieved by electric motors powered by electrical generating means carried by the platform or by electric accumulators. The use of electric accumulators to power electric motors and electric actuators for Movements of the mobile platform 11 and the turret 12 is suitable for inspection of aircraft that is performed in an airport environment with easily infrastructure to allow recharging of the electric accumulators between two inspections or as much as necessary.

The collaborative robot further comprises calculation means 20 necessary for visual inspection.

 The calculation means 20 comprise in particular acquisition means 21 for the observation data transmitted by the display means 13, processing means 22 for the said observed data, communication means 23 with a control center 50, deported by report to an area in which the collaborative robot is to perform inspections, and location means 24.

 The various means 21, 22, 23, 24 of the calculation means 20 on board the robot 10 are advantageously means communicating with each other digitally via a communication bus of said robot.

 The acquisition means 21 of the observation data mainly consist of digital data storage memories having sufficient capacity to keep the observation data for a period necessary for their processing in order to determine whether anomalies have been observed and if need of data preprocessing means for their exploitation.

 To the extent that comparative data may be necessary for the inspection of a given aircraft, for example between two zones of the aircraft, advantageously all the observation data acquired during a given inspection are kept at least during the inspection period.

In one embodiment, the observation data are stored in memory by the robot 10 or in a remote storage center, not shown, for subsequent downloading by the robot, with reference to the aircraft on which said observation data has been established, so that the comparative processing of the observation data between two or more inspections can be performed. The processing means 22 mainly comprise digital processing units, for example processors and or microcontrollers, random access memories (RAM), storage memories (ROMs, flash memories, SDRAMs, hard disks, etc.), interfaces digital or analog and communication buses between the different components of said processing means.

 As far as necessary, the processing means 22 comprise specialized circuits for the rapid processing of the signals transmitted by the display means, for example circuits for the processing of the signals corresponding to fixed or moving images.

The processing means 22 are configured to produce sequences of instructions ensuring:

 configuring the robot 10 according to operations to be performed, in particular an initialization according to the aircraft to be inspected visually;

 communications with distant means;

a pilot implementation resources to the movements of the platform 11, in particular the control of the propulsion engine of the platform and management of energy in the flat form ^¬ resources;

 control of the actuators of the turret 12 and or viewing means 13;

 the processing of observation data;

 - the development of diagnoses based on the observation data processed;

 a precise location of the robot 10 and a point observed on an outer surface of the aircraft 90 being inspected, in a reference linked to the aircraft, if not continuously at least during observations to give rise to a anomaly finding.

In general, the robot comprises the mechanical means and which appear to those skilled in the art necessary for the implementation described as an example below.

As will be apparent from the exemplary implementation described in detail, the collaborative robot of the invention is not a fully automated means of inspection for detecting anomalies on an aircraft 90 but a means of assistance available to an operator for performing a visual inspection.

 The collaborative robot is particularly adapted to allow remote inspection by the operator and to provide inspection assistance for more accurate, detailed, repeatable and reliable diagnoses.

According to the visual inspection method 200, in a first step 210, when an aircraft 90 is to be inspected, for example for a pre-flight survey, a collaborative robot is initialized to be able to perform the desired inspection. .

 During this first step 210, the robot 10 receives a position 211 of the aircraft 90 to be inspected, at least one approximate position such as the number or the geographical position of a block on which the aircraft is placed, and receives the information 212 on the aircraft to be inspected, at least one type of aircraft and preferably also an identity of the aircraft.

 It should be noted with regard to the position of the aircraft that said position may be implicit, for example in the case where the collaborative robot is assigned to a particular block and that in this case the aircraft will still be, for the robot , on the same site.

 When, on the other hand, the collaborative robot may be required to carry out inspections on aircraft that can be parked on different locations of a parking area, it will then be informed during the first step 210 of the location where the robot is located. 90 aircraft to inspect.

In an implementation mode during a second step 220, the collaborative robot 10, when it reaches the area in which the aircraft 90 which it is to inspect, examines the aircraft overall with its means. 13 from which examination it deduces, from a database of aircraft shapes, the type of aircraft in question. Preferably, the collaborative robot, by processing the images acquired during this step, detects the registration of the aircraft and by consulting a database of registered aircraft verifies that the type of aircraft bearing said registration corresponds to the type identified by the analysis of the shape of the aircraft.

 In one embodiment, the information on the aircraft 90 obtained during the first step 210 and during the second step 220 are compared during a third step 230 to check the consistency of the data transmitted by the control center. to detect any inconsistency as soon as possible.

 Any anomaly or inconsistency detected during the third step is immediately reported by the collaborative robot to an operator responsible for the inspection who will decide on the arrangements to be made: stop the inspection by the robot in order to correct information on the type and identify the aircraft, send the robot orders to obtain additional visual data before making a decision, give the collaborative robot instructions to continue the inspection by ignoring certain inconsistencies detected, if necessary by also requesting a intervention by an operator to carry out an on-site analysis and remove the doubt.

In a fourth step 240, which fourth step can be carried out in the previous steps on the basis of the instructions received by the collaborative robot, said collaborative robot downloads the data necessary for the inspection to be performed, in the case at least where it does not have all the said data in its internal storage memories.

Said data comprise, for example, an inspection circuit to be carried out determined by a nominal trajectory, lists of operations to be performed under the aircraft type inspection, lists of inspection operations to be performed. for the particular aircraft, for example, based on previous observations on the same or other aircraft of the same type In one form of implementation, said data, at least for some of them, are acquired by the collaborative robot by learning, for example during inspections carried out in remote-controlled mode, and are completed during successive inspections to improve future detection of anomalies that have been detected.

 The said data also include the characteristics of the aircraft, nominal geometrical characteristics, colors and patterns painted on the aircraft, positions and types of hatches, locks, probes ... to be verified during the visual inspection, anomalies known, for example detected during previous inspections ...

In a fifth step 250, the collaborative robot will perform the visual inspection operations following the inspection circuit while moving on the ground near the aircraft 90.

 During its movements, the collaborative robot will continuously determine its position in a reference frame linked to the aircraft being inspected, for example an OX-OY landmark whose trace corresponds to a longitudinal X axis of the aircraft. the aircraft and a transverse Y axis of the aircraft.

 In order to determine its position, the collaborative robot uses conventional techniques, for example the recognition of bitters constituted by characteristic forms or subassemblies of the aircraft, data which are advantageously coupled to geolocation information, for example by GPS. , by laser telemetry or other localization devices, and odometry techniques to integrate the movements of said robot, the various techniques being advantageously hybridized to make accurate estimates of the position and orientation of the robot in the OX-mark. OY, typically about one centimeter or less in position and one minute or less in orientation.

Although the collaborative robot 10 is called upon to operate on a substantially flat and horizontal surface of an aircraft parking area of an airport facility, where appropriate the location techniques used ensure the determination of a relative position. in height of the robot by report to the aircraft.

 If necessary, it is also determined the inclination of the platform with respect to a local vertical, in particular if said collaborative robot is not provided with a stabilizing function to keep the mobile platform 11 horizontal, or an axis of the turret 12 vertical.

 The robot 10, following the instructions stored in a program of the processing means, will move along a path allowing it to visualize all the points of the aircraft to be inspected.

 The robot 10 generally follows a predetermined trajectory 91 on the ground.

 During its movements, the robot 253 detects any obstacles, for example by optical means and or by specific sensors, for example by ultrasonic emission sensors and where appropriate deviates from the predetermined trajectory. performing the necessary movements to circumvent the obstacle (s) and to ensure that visual inspection of all areas to be inspected is properly performed.

 In addition, a detected obstacle is reported 254 to the operator responsible for the inspection as a potential anomaly that must be addressed before the aircraft 90 leaves the parking spot.

 To perform the visual inspection, the robot according to its position and its estimated orientation in the OX-OY mark positions and directs the display means so that images of the zones of the aircraft to be inspected are taken and transmitted. to the processing means 22.

 In one embodiment, the images are made continuously during the movements of the robot and its stops so as to visualize areas of the aircraft that can be observed from the ground given the possibilities of orientation and elevation over the soil of the visualization means 13.

The processing means 22 perform a processing of the received images 256 to identify any type of anomaly resulting in identifiable characteristics in the images. For example, such characteristics correspond to geometrical and or contrasts and or colorimetric singularities, singularities that can be observed in one or in different illumination spectra, the robot switches the illumination spectra during the positioning and orientation step 255.

 In one operating mode, the robot 10 selects a given spectrum of illumination means 132 according to the types of anomalies sought in an area being inspected or according to the facies of a detected anomaly, for example in white light. , to establish a diagnosis.

 The spectrum of illumination is for example in the field of visible light, in white light or in a particular color of the visible spectrum, or in the infrared or ultraviolet range, for example to identify areas having particular fluorescences.

 Such fluorescence properties may be the direct trace of an anomaly such as, for example, leakage of a liquid, such as fuel, hydraulic fluid or the like, leaving a visible trace on an outer surface of the aircraft and or on the ground. They may be for example the sign of an aircraft electrically charged by electrostatic charges, by the presence of corona discharges perceptible at static fuselage unloaders or wings.

 They can also be the result of markers revealed by particular conditions for example following exposure to abnormal heat or shock. Such markers consist, for example, of paints whose spectral properties are modified by exposure to heat or by paints comprising microspheres filled with a fluid that becomes visible when illuminated by wavelength light. particular and whose fluid is dispersed during a mechanical impact on a structure covered with such paint.

In the case of visible structural parts, the image processing is carried out, in addition to the detection by developers such as those mentioned above, to detect abnormal shapes or changes of shape between successive inspections, or dissymmetries of forms between nominally symmetrical areas of the aircraft, for example bulges or depressions, such changes of shapes being interpreted in general as anomalies of the structure. In the case of the particular areas to be inspected to verify their status and integrity, the processing means 22 perform an analysis of the images transmitted by the display means 13 by analogy with virtual representations of the observed areas, in particular of elements located on these particular areas, kept in memory. The virtual representations are, for example, three-dimensional views or digital representations that make it possible to place a virtual representation in the position, orientation and distance, under which the actual particular zone is observed by the visualization means 13 in order to carry out a treatment. digital comparison of the real element with its virtual representation. In this particular case, anomalies of greater or lesser dimensions can be sought according to criteria established in advance, the display means 13 being advantageously provided with magnification means, for example of the optical zoom type, to perform the search for small size defect.

 The particular zones are for example zones comprising a door or a inspection hatch, or engine hoods whose closure and the position of the locking devices must be verified.

 The particular zones are, for example, zones comprising visible equipment such as an incidence probe, a total aerodynamic pressure tap, a static pressure tap, a frost detector, an antenna, a drain or any other type of can be damaged, for example twisted or torn off, or closed, for example by pit caches During the inspection, the collaborative robot transmits to the command center 50 supervised by an officer of the inspection of the aircraft reports inspection.

If the robot 10 does not detect any anomaly 258 in an inspected area, it transmits information of absence of detected anomaly, for example being translated in the control center by a "green light" and information of progress of the inspection can be displayed on a screen 51 of the control center 50 to be followed. If the robot detects an anomaly 259, it retransmits an alert to the command center 50, which alert is accompanied by images of the visual observation of the area that occurred at the detection of the anomaly with a diagnosis or a list of diagnostics possible according to their probabilities given the typology of the anomaly treated by the robot 10.

 In a case where the collaborative robot identifies an anomaly associated with a high probability that the anomaly is affecting the security, said collaborative robot interrupts the inspection and transmits to the control center 50 an abnormality detection report with the message. warning and a warning "red warning", an intervention on the aircraft being in this case necessary a priori.

 In the case where the collaborative robot 10 identifies an anomaly for which said collaborative robot is not able to make a diagnosis, if not with a high uncertainty on the consequences of the anomaly, said collaborative robot transmits the alert message with an "orange" alarm (caution) and goes into a standby condition and in a remote mode 262 in which he waits for instructions from the operator responsible for the inspection.

 The operator can, after examining the images and data transmitted by the collaborative robot, decide either:

 - Send the robot the order to continue the visual inspection 263 considering that there is a false alarm or that the anomaly detected is minor;

 - Remote control 264 the robot to obtain additional data on the area that generated the alarm and can make a decision to continue 263 or stop 265 the visual inspection;

 - Complete the inspection performed by the robot 10 by an intervention on the aircraft 90 of an inspection team.

In a particular form of implementation, the data collected by the robot 10 and the diagnostic information provided by the operator via the control center 50 are processed to enrich fault databases and to ensure subsequent inspections. better detection of the type of anomaly encountered by the robot 10 performing the visual inspection or by a fleet of inspection robots.

It should be noted the interest for the robot 10 to take into account all the specificities of the aircraft 90 it inspects. It is indeed not exceptional that a given aircraft has peculiarities compared to other aircraft of the same type.

 For example, a given aircraft may comprise an antenna missing on other aircraft of the same type or an antenna of a different model, a situation that may lead to the detection of a false anomaly.

 The knowledge by the robot 10 of the exact identity of the aircraft 90 allows it to have information specific to this aircraft and for example the presence in a zone of a particularity of the aircraft in question, for example an antenna attached to this particular aircraft.

 Furthermore, if the database of the characteristics of the aircraft 90 does not include the presence of an observed characteristic, the operator responsible for the inspection, notified by the robot 10 of a detected anomaly, will be able to see the appearance normal for example of the antenna and confirm this information to the robot that will enrich the database for the aircraft 90 for subsequent inspections and thus avoid new alerts for the detection of the same false anomaly.

When the robot 10 searches during an inspection of possible anomalies on an aircraft 90, during the fifth step of the method, it inspects successively the different zones of the aircraft to detect visible traces attributable to an abnormal state of the aircraft. 'aircraft.

 For this, the robot 10 moves around the aircraft 90 by methodically scanning with the display means 13 the different parts of the aircraft, or even the ground under the aircraft. The images obtained are, after treatment, compared with data representative of normal situations and abnormal situations to detect the presence of a possible anomaly.

For the purposes of this detection, the images obtained by the visualization means are as much as necessary processed in order to be compared to the data of the known database of the robot 10.

 The images are also analyzed to identify any generic characteristics of potential defects such as deformations on the outer surface of the aircraft 90, missing parts resulting in practice by irregular openings or locations for which it does not exist. is not known to the robot the presence of opening, traces such as scratches or drips on the outer surface of the aircraft, lack of paint in some locations, traces with a facet of lightning impact ...

 The defects potentially associated with these features are not in practice sought at a particular location but may be at many points on the outer surface of the aircraft.

 The processing applied to the images may be general or specific to the detection of a particular problem. For example, the images are processed by contrast enhancement algorithms, extraction of image areas according to its color or its spectral sensitivity, contour extraction, texture ...

 The detection may for example relate to a shape or a color that does not conform to the shape or the nominal color, the shape or color observed may reflect a deformation or the presence of a foreign body.

 An anomaly of form results in an observed form that does not correspond to the expected form.

 When an anomaly is detected, the robot 10 characterizes the anomaly on the one hand by its type, that is to say the characteristics which led to consider the presence of an anomaly, for example: the shape, the color, range, sensitivity at certain wavelengths, presence or absence of an unexpected or expected element, contrast ... and secondly quantifies the anomaly.

The robot also locates the anomaly observed in the axis system of the aircraft, location that it realizes taking into account its precise position in the axis system Ox-Oy in which it operates on the ground, the direction in which are oriented the display means and the geometric characteristics of the observed aircraft.

 The knowledge of the position of the anomaly on the outer surface of the aircraft is advantageously exploited by the processing means 22 to perform a diagnosis of the anomaly. For example, it is known that lightning attaches to the privileged locations of an aircraft and the presence of a trace having a lightning impact facies will not necessarily be interpreted in the same way as to its causes and effects. according to the location where the trace is.

 To take the example of a lightning impact, the knowledge of the position of the anomaly on the surface of the aircraft also makes it possible to relate the anomaly to structural characteristics of the aircraft at the location of the aircraft. 'anomaly. For example, the structure may be a metal structure or a composite material structure with properties depending on the type of composite material, and in the latter case with a protection against lightning currents which may be different, for example a wire mesh. more or less dense bronze or a particular conductive paint.

 The knowledge by the robot 10 of the physical and dimensional characteristics of the surface of the aircraft 90 in which an anomaly of a particular type is identified, as well as characteristics of the structure not visible under the visible surface, for example the presence of a smooth or structural frame, a support equipment or equipment, the processing means make a prognosis so as to provide the operator in charge of the inspection detailed information on the anomaly and the consequent risks.

 The detection of a shock type anomaly on the outer surface of the aircraft during the inspection results, for example, in a message intended for the control center of the type "fuselage coating depression at the level of the beam No. 14 between the fuselage frames 11 and 12 - no traces of coating tear - low structural risk, repair to be programmed "accompanied by images of the area concerned.

In a preferred form, the processing means perform a quantitative analysis of the defect: extent (surface, length, width, etc.) of the anomaly, by example of a surface affected by a shock, depth of a deformation, intensity of a cause that must have led to the observed anomaly such as a rise in temperature, intensity of a leak in the case of a presence of suspicious fluid ...

 To perform this quantitative analysis, the robot 10 uses as much as possible its visual means associated with the processing means as already specified by stereoscopic vision techniques and or by any known optical method such shearography, holography, telemetry .... In a mode implementation, the robot performs structural examinations by means of non-destructive testing instruments to quantify the anomalies.

 When an anomaly is detected in step 259 by the visual inspection, the robot according to the diagnosis made is positioned, in step 2591, with respect to the anomaly to deploy the non-destructive testing instruments, not shown, to obtain additional information to improve diagnosis based on visual inspection.

 Such non-destructive control means can implement probes, for example ultrasonic probes, eddy current probes, temperature measuring means such as thermal imaging cameras or any other type of probe capable of producing a local examination of the material in a particular area.

 Advantageously, the probes are carried by one or more articulated arms so that the robot can apply / orient said probes on / to the locations to be examined, each probe being, during an examination using said probe, associated with an equipment of measurement embedded on the platform of said robot.

 In one embodiment, in which the robot comprises a plurality of probes, the probes are stored in a probe holder of the robot to be implemented via a single articulated arm.

To perform the nondestructive examination, the articulated arm, controlled by the robot processing means or by a control center operator 50 having chosen to remotely control the examination, enter the appropriate probe for a given examination, by any type of gripping and connecting means to the measuring instrument to which the probe is to be connected, then apply / orient the probe on / to the zone to be controlled by ensuring the desired displacements for said probe, and when the control has been carried out and the processing means and / or the remote operator consider having obtained the data necessary for the inspection, the articulated arm replaces the probe in the probe holder.

 According to the results of the examination transmitted to the control center 50, if necessary with a diagnosis made by the processing means 22, and subject to instructions received from said control center, the robot 10 continues the inspection.

 When the inspection of the aircraft or the part of the aircraft entrusted to the robot 10 is completed, said robot returns to a waiting station 30 during a sixth step 270, advantageously a waiting station providing protection for said robot to prevent damage by the various vehicles and aircraft traveling on the airport platform, and preferably a position where said robot is connected to a power source allowing him for example to recharge his accumulator batteries electric if provided.

 In one embodiment, the waiting station 30 is provided with communication means 31 at short distances with the robot, where appropriate wired connections, to allow rapid exchange of information and with little risk of interference between the control center 50, for example to transmit all the data collected during the inspection in view of a storage of said data and a possible deferred processing.

 Such means may also be implemented in the first step 210 when the robot 10 loads the useful information for its next inspection.

In one embodiment and implementation, the data transmitted by the robot 10 to the control center 50 during the visual inspection may be received by cockpit equipment so that a member of the technical crew has the opportunity to follow the progress of the inspection and to assess any anomalies detected.

 In a variant of this embodiment, control means of the robot are arranged in the cockpit to intervene on the progress of the inspection.

Advantageously, despite the possibility of remote intervention on the robot during the visual inspection, the robot resumes the inspection so that all the requirements nominally provided for an inspection are fulfilled before the inspection can be declared complete.

 In this case, a "light" showing the status of the inspection remains "red" until all the requirements have been fulfilled, if necessary "orange" if only requirements deemed incidental by the person responsible for the inspection. have not been realized.

During his movements, both during the inspection and to get to the location of the inspection to leave the said location, the robot 10 provides security to avoid collisions with persons or objects that may be fixed or animated. in its close environment. For example, the robot 10 uses its display means 13 to detect the objects with which a collision can occur. The robot may also include means specifically dedicated to the detection of such objects.

When the processing means 22 identify or are warned of a risk of collision, said means control the movements of the platform ^¬ to avoid the collision, if necessary by emitting a sound signal and or light or other to alert a person who can to be involved in the potential collision.

Advantageously, the processing means provide guidance of the platform autonomously to avoid the risk of collision by using known localization techniques with respect to an environment whose characteristics they reconstruct, for example using a SLAM method. (Simultaneous Localization and Map Building).

According to the embodiment as just described, a single robot 10 performs the visual inspection of the aircraft 90.

 In another embodiment of the visual inspection method, the robot 10 performs partial visual inspections in combination with partial inspections performed by other similar or specialized robots.

 For example, several robots can be implemented, not shown solution, to inspect different parts of the aircraft so that a complete inspection can be performed in a reduced time.

 In this case, for example, two or more identical robots are used to carry out the visual inspection of a right-hand side of the aircraft and of a left-hand side of the aircraft, and / or of a front part. and a back part.

 In this case the inspections carried out by two robots are for example synchronized in order, for example, to obtain comparative data between two zones of the aircraft such as a zone on the right side and a theoretically symmetrical zone on the left side so as to identify or confirm the presence of an anomaly.

 It can also be implemented robots with different possibilities, for example adapted to parts of the aircraft for which it is necessary to implement particular viewing means, for example because of their heights, such as a drift of an aircraft, or their shapes, such as air inlets or outlets of engine nozzles. In a particular embodiment implementing non-destructive control means, all or part of the non-destructive control means are carried by at least one control robot whose mobile platform is independent of the visual inspection robot.

In this case, the control robot is controlled by the visual inspection robot which, when it has identified an area to be subjected to non-destructive testing, sends an intervention command to the control robot with all the data necessary for the control robot to perform the requested control, in particular the exact location on the aircraft of the area to be controlled and its extent and the type of control desired. When the control robot has performed the requested control it transmits to the visual inspection robot the data obtained by the control, possibly processed to provide an interpretation of the control, and said control robot is placed in a waiting position.

 The implementation of a separate control robot requires the realization of a specific mobile platform for this control robot, but makes it possible to simplify the visual inspection robot, to reduce its cost and to increase it. 'agility.

 In addition, the same control robot, which is intended to intervene only occasionally and at the request of a visual inspection robot, can serve several visual inspection robots, priority rules managing potential conflicts if two robots of visual inspection solicit the control robot at the same time.

 In one embodiment, different types of control robots are implemented, for example a robot for performing ultrasonic checks and another robot for performing eddy current checks. It is in this form obtained more or less specialized control robots whose performance can be adapted to each type of control.

 The example of visual inspection robot 10 described in the exemplary embodiment is a robot moving on the ground, however other types of robots, for example levitating robots, can be implemented if necessary in combination with one or more robots on the ground.

 In one embodiment of the robot 10, said robot comprises means, for example carried by an articulated arm, for opening and closing doors on the surface of the aircraft giving access from outside the aircraft 90 to outlets. or indicators so that the robot 10 is able to close a door left open by mistake, or to open a door to visually inspect a housing and close the housing door.

It is thus possible to realize with a visual inspection robot 10, or with a reduced number of inspection robots, a visual inspection of an aircraft, for example an airplane, in a non-specific environment, for example on a parking space of an airport.

 The inspection is carried out collaboratively between the robot 10 and the remote operator responsible for the inspection to which said robot returns all the results of the visual inspection, if necessary non-destructive test results of an area presenting a visually detected anomaly, from which he waits for instructions whenever an anomaly is detected and a decision requires the intervention of the operator to whom the images are presented and all the characteristics that could have been measured of the anomaly .

 Depending on the device and the visual inspection method, the operator responsible for the visual inspection remains totally in control of the decision to declare the absence of any anomaly affecting the operational use of the inspected aircraft with the possibility of intervening remotely on the robot to obtain all the information, in particular the images of dubious zones, to make a fast decision adapted to the case.

 The control center from which the person responsible for the inspection operates may be distant from the place of the visual inspection, for example a visual inspection officer may be located in a maintenance center of the aircraft operator and supervise visual inspections at airports around the world, in particular by making use of possible digital links over landlines, terrestrial radio networks and satellite radio networks.

Claims

Apparatus for visually inspecting (100) the outer surfaces of an aircraft (90) having an inspection area for receiving an aircraft characterized in that said visual inspection device comprises at least one inspection robot (10) visual device with a mobile platform (11) carrying a turret (12) with display means (13) and comprises processing means (20) providing a guidance of the mobile platform (11) and a processing of the information received from the display means (13), characterized in that said visual inspection device (100) comprises a control center (50) with a station for at least one control operator, and characterized in that the processing means (22) of the visual inspection robot are suitable for:

 driving the visual inspection robot (10) autonomously during a visual inspection of the exterior surfaces of an aircraft (90) parked on the inspection area;

 discontinue a visual inspection in progress if an anomaly is detected on the outer surface of an aircraft being inspected;

 transmit visual inspection information to the control center;

 receive instructions from the control center for the follow-up to be given to a visual inspection.

A visual inspection device according to claim 1 wherein the visual inspection robot comprises means for determining at any time during an inspection a position of said visual inspection robot and an orientation of the viewing means in a viewing system. axis related to the aircraft.

A visual inspection device according to claim 2 wherein the robot processing means is adapted to determine the position of the robot and the orientation of the display means by an image processing of the aircraft to be inspected obtained by the display means.

Visual inspection device according to claim 2 or claim 3 wherein the visual inspection robot comprises absolute locating means, such as GPS receiver, optometry or laser rangefinders pointing reference targets, and or integration means of his displacements.

Visual inspection device according to one of the preceding claims wherein the processing means comprise data storage means comprising, at least temporarily, characteristics, in particular geometric and graphic, of an aircraft to be inspected. 6 - visual inspection device according to one of the preceding claims wherein the processing means comprise data storage means having abnormality characteristics, for example in an anomaly library. 7 - visual inspection device according to one of the preceding claims wherein the processing means comprise image processing algorithms for detecting on images transmitted by the display means abnormalities visible by said display means in a wavelength of the optical spectrum.

8 - visual inspection device according to one of the preceding claims wherein the display means comprise means for illuminating in a light of the visible range and or in the field of the infrared and or in the field of ultraviolet .

9 - visual inspection device according to one of the preceding claims wherein the display means and the processing means are configured to determine a three-dimensional shape of the inspected outer surfaces of the aircraft. - Visual inspection device according to one of the preceding claims comprising non-destructive testing means of the structure of the inspected aircraft. - Visual inspection device according to claim 10 wherein all or part of the non-destructive control means are carried by the visual inspection robot. - A visual inspection device according to claim 10 or claim 11 wherein all or part of the non-destructive control means are carried by at least one control robot whose behavior is controlled by the visual inspection robot. - Visual inspection device according to one of the preceding claims wherein the viewing means are orientable in elevation and azimuth relative to a reference platform of the visual inspection robot. - visual inspection device according to one of the preceding claims wherein the processing means are configured to determine a position of a fault detected on an aircraft with respect to elements of the internal structure of said aircraft, not visible since the outside the aircraft. - visual inspection device according to one of the preceding claims wherein the inspection robot is a robot rolling on a floor of the inspection area or moving by levitation in a volume whose footprint at soil corresponds substantially to the inspection area. 16 - visual inspection device according to one of the preceding claims comprising a plurality of inspection robots configured to jointly perform a visual inspection of the same aircraft. 17- A method of visual inspection of an aircraft in which images of an outer surface of the aircraft to be inspected are transmitted to processing means of a visual inspection robot, wherein the processing means analyzes the images for identifying the presence of possible visible anomalies, characterized in that when a visible anomaly is detected, the data relating to the detected anomaly is transmitted to a control center and in that the visual inspection is interrupted, at less when the processing means identify an anomaly belonging to a category of anomalies considered critical. 18 - visual inspection method according to claim 17 wherein when the inspection is interrupted due to the detection of an anomaly, instructions are transmitted to the visual inspection robots by the control center for the pursuit of the visual inspection, said instructions determining how the robot should continue the inspection.

19 - Visual inspection method according to claim 17 or claim

18 in which an amplitude of a visible anomaly is calculated by the processing means from optical deformation measurement means and or by a colorimetric analysis in the visible range, and or infrared, and or ultraviolet light spectrum.

20 - visual inspection method according to one of claims 17 to 19 wherein when an abnormality is detected visually, an area affected by the visible anomaly is subjected to a non-destructive inspection by the visual inspection robot or by a non-destructive control robot controlled by the visual inspection robot.