METHOD FOR CREATING AN AUTONOMOUS NAVIGATION MAP FOR A VEHICLE

Description

[0001] The present invention relates to a method for developing a navigation autonomy map for a vehicle, the autonomy map obtained and an application of this map to determining a route. BACKGROUND OF THE INVENTION

[0002] Conventional vehicles today are widely equipped with a navigation system that includes map data and implements a route planning algorithm that uses this data. These systems, commonly called GPS (abbreviation for Global Positioning System), are not, however, designed to allow driverless vehicle operation on their own.

[0003] However, we know that significant work is being carried out with the aim of enabling autonomous road navigation, i.e. without the intervention of a pilot, for vehicles circulating on the road network.

[0004] The emergence of autonomous functions in vehicles raises the issue of the risk of failure of these functions and their consequences. These functions only provide navigation / vehicle control solutions if all the necessary external information is present and properly processed by the onboard system. Otherwise, they either fail and the driver must take over, or they make a mistake, and taking over by the driver is even more difficult.

[0005] There are currently various ways to ensure a vehicle's autonomous navigation function or at least participate in this autonomous function. The main techniques currently available are known as: GPS (already mentioned above); LIDAR, acronym for Light Detection And Ranging, or LADAR, acronym for Laser Detection And Ranging; INS, abbreviated from the English expression Inertial Navigation System, or in French CIN for Centrale Inertielle de Navigation; techniques using the detection of road markings; techniques exploiting the emission of dedicated signals by transmitters placed in road infrastructure equipment, in particular traffic lights.

[0006] WO 2014 / 139821 A1 discloses that a final autonomy index (PA) is calculated from a simple sum of the primary autonomy indices calculated for each journey segment according to at least two distinct navigation autonomy functions.

[0007] None of these techniques is satisfactory for covering an entire road network on its own. SUBJECT OF THE INVENTION

[0008] An aim of the invention is to propose a method for developing a navigation autonomy map for a vehicle, which is reliable over all sections of the journey in a given area. SUMMARY OF THE INVENTION

[0009] In order to achieve this goal, the invention proposes a method for developing a navigation autonomy map for an autonomous vehicle, according to claim 1.

[0010] Thus, by choosing complementary autonomy functions adapted to the configuration of the area considered, we obtain for each section of the journey a final autonomy index presenting improved reliability compared to an existing technique. The autonomy map thus groups together specific cartographic data such as the autonomy indices which will make it possible to increase the safety of autonomous vehicles, for example by making it possible to plan journeys whose sections are assigned the autonomy indices most favorable to autonomy.

[0011] According to an advantageous version of the invention, the final autonomy index is calculated by performing a weighted average of the primary autonomy indices and the method comprises the steps of: performing several calculations of each final index with distinct weighting coefficients; comparing the final autonomy indices obtained; and developing the navigation autonomy map by retaining the weighting for which the lowest final index is the strongest of the lowest final indices of the different weightings.

[0012] This further improves the accuracy of the resulting navigation autonomy map.

[0013] According to another advantageous version of the invention, the method comprises the steps of: performing several calculations of each final index with distinct weighting coefficients; comparing the final autonomy indices obtained; and developing the navigation autonomy map by retaining the highest final autonomy index for each section of the journey.

[0014] According to another aspect of the invention, the method comprises the steps of: identifying journey sections for which the final autonomy indices are lower than a predetermined threshold, and performing smoothing on the immediately adjacent journey sections. Preferably, the method comprises the prior step of analyzing the configuration of the journey sections and installing equipment suitable for improving the primary autonomy indices.

[0015] The invention also relates to: a navigation autonomy map for an autonomous vehicle in an area comprising journey sections, in which, for each journey section, the map comprises a final autonomy index obtained by performing a weighted average of primary autonomy indices according to at least two distinct autonomy functions; and an application of this autonomy map to the determination of a journey obtained by adding autonomy parameters to the different parameters (speed, existence of tolls, etc.) which are currently used to determine a journey. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Other characteristics and advantages of the invention will become apparent upon reading the following description of a preferred non-limiting embodiment of the invention with reference to the attached figures, among which: There figure 1 is a block diagram of the method according to the invention, The figure 2 is a map comprising primary indices calculated for a first autonomy function, The figure 3 is a map with primary indices calculated for a second autonomy function, The figure 4 is a map with primary indices calculated for a third autonomy function, The figure 5 is a map with final indices, calculated from the primary indices of the previous figures. DETAILED DESCRIPTION OF THE INVENTION

[0017] With reference to the figures, the illustrated process comprises a first step 1 of analysis of the configuration of the route sections, followed by a choice 2 of the autonomy functions. For example, for a city with a very closed structure, comprising very tightly packed buildings, and heavily equipped with ground markings and dedicated transmitters placed in city facilities, the corresponding recognition parameters will be chosen, while on a road in open countryside the GPS autonomy function will be chosen.

[0018] We then perform calculation 3 of the primary autonomy indices, followed by calculation 4 of the final indices by performing a weighted average of the primary indices. The weighting coefficients correspond to the relative importance of each primary index for the autonomy of the vehicle (for example, in a given section, having precise road markings and high-definition mapping may be more important than having a visual reference point in line of sight). The weighting coefficients can be determined a priori or be updated periodically or in real time according to traffic conditions (climatic event such as fog, presence of an accident or roadworks).The weighted average can be a constrained weighted average (in this case, it is considered that when the primary index is less than a value X, the coefficient associated with it is worth 0, while when the primary index is greater than the value X, the coefficient is worth 1). For example, imagining that the law prohibits autonomous circulation when the marking is erased, the use of the constrained weighted average would make the criterion relating to the road marking preponderant over the others when the road marking is absent.

[0019] If the final indices obtained do not appear sufficiently reliable, a new calculation of the final indices is preferably carried out with a different weighting, followed by a comparison 5 and a selection 6, followed where appropriate by smoothing 7, for example using an inertial navigation unit.

[0020] There figure 2 illustrates the result of a calculation of primary indices on a district of Paris using the GPS autonomy function. The value of the primary index is illustrated for four decreasing values ​​by the following signs: + sign, thin hatching, thick hatching, small circles.

[0021] There figure 3 illustrates the result of a calculation with the LIDAR autonomy function, using the same signs for the representation of the indices obtained.

[0022] There figure 4 illustrates the result of a calculation with the ground marking autonomy function, always using the same signs.

[0023] There figure 5illustrates the result of a weighted average calculation from the three previous ones. It can be seen that the quality of the GPS index on a part of the rue de Rivoli is partially deteriorated by the other autonomy functions. A correction can however be made with a CIN autonomy function. It should be noted in this regard that the quality of the index depends on the quality of the components used. In particular, a good quality CIN would make it possible to move all sections of the journey to the higher value. The quality of the components combined with the method of the invention will therefore make it possible to distinguish between vehicles truly equipped for autonomous navigation, and those which are only autonomous in specific circumstances.

[0024] The autonomous vehicle enabling the implementation of the method of the invention comprises an automatic piloting unit and a navigation unit. The automatic piloting unit is connected to the vehicle's control members such as the steering system, the engine, the braking system, as well as to sensors for detecting the vehicle's behavior and the vehicle's environment, and incorporates calculation means for controlling the control members based on the information provided by the sensors so that the vehicle remains on a determined trajectory while complying with road regulations. The navigation unit comprises, for example, a GPS module and a route planning module based on cartographic data including the autonomy map.The autonomy map is used both by the navigation unit to determine a route according to the index of the sections taken and by the control unit which will determine, according to the section taken, which sensors it will prioritize (for example, in an area in which the reception of GPS signals is of poor quality, the GPS information will be neglected for the control of the vehicle so as to limit the risk that taking this information into account will alter the precision of autonomous control).

[0025] Autonomy maps can be used to advantage by an organization managing the road network to identify sections of its network that need to be improved for the circulation of autonomous vehicles (by redoing road signs, installing RFID tags to provide information to autonomous vehicles, etc.).

[0026] Autonomy maps can be used to advantage by car manufacturers to help them determine how much of the network their vehicles can operate autonomously on and to improve their vehicles by helping them identify which sensors should be fitted to autonomous vehicles to enable autonomous driving on a larger part of the network.

[0027] Of course, the invention is not limited to the embodiment described and variant embodiments can be made without departing from the scope of the invention as defined by the claims.

[0028] In particular, it is understood that the autonomy of a vehicle depends on the sensors present on the vehicle. Consequently, an autonomy map will be developed for each model, or even version, of a vehicle. It will also be possible to provide organizations managing the road network at the local or national level with "global" autonomy maps corresponding to several types or models of autonomous vehicles. For example, a "minimum" autonomy map could be established based on the minimum equipment of autonomous vehicles and a "maximum" autonomy map based on the maximum equipment of autonomous vehicles.

Claims

1. A method of preparing a navigation autonomy map for a autonomous vehicle equipped with at least two sensors each providing a distinct navigation autonomy function, covering a zone including path segments, the method comprising the following steps: identifying (1) the path segments for which the map needs to be prepared; for each path segment, calculating (3) primary autonomy indices for each of navigation autonomy function; and calculating (4) for each path segment a final autonomy index on the basis of the primary autonomy indices, characterized in that the final autonomy index is calculated by taking a weighted average of the primary autonomy indices according to a relative importance of each primary autonomy index for the navigation autonomy of the vehicle.

2. A method according to claim 1, including the following steps: performing a plurality of calculations (4) for each final index with distinct weighting coefficients; comparing (5) the resulting final autonomy indices; and preparing (6) the navigation autonomy map by retaining the weighting for which the lowest final index is the highest of the lowest final indices of the various weightings.

3. A method according to claim 1, including the following steps: performing a plurality of calculations (4) for each final index with distinct weighting coefficients; comparing (5) the resulting final autonomy indices; and preparing the navigation autonomy map by retaining the highest final autonomy index for each path segment.

4. A method according to claim 1, characterized in that it includes the steps of identifying path segments for which the final autonomy indices are less than a predetermined threshold, and of performing a smoothing operation (7) applied to the immediately adjacent path segments.

5. A method according to claim 1, characterized in that it includes a prior step of analyzing (1) the configuration of the path segments and of installing equipment adapted to improve the primary autonomy indices.

6. A navigation autonomy map, resulting from the method according to claim 1, for a vehicle covering a zone comprising path segments, the map being characterized in that for each path segment it includes a final autonomy index obtained by taking a weighted mean of primary autonomy indices in accordance with at least two distinct autonomy functions.

7. The application of an autonomy map according to claim 6 for determining a path, characterized in that autonomy parameters are added to geographical parameters for determining the path.

8. An autonomous vehicle including an automatic driver unit and a navigation unit; the automatic driver unit being connected to vehicle control members and to sensors for detecting the behavior of the vehicle and the surroundings of the vehicle, and incorporating calculation means enabling the control members to be driven on the information supplied by the sensors so that the vehicle remains on a determined path in compliance with traffic regulations; the navigation unit comprising a module for planning itineraries on the basis of map data comprising an autonomy map according to claim 6, the autonomy map being used both by the navigation unit to determine an itinerary as a function of the indices of the segments traveled, and also by the driver unit to determine which sensors should be given precedence on the basis of the segment traveled.