Method for recognizing a construction site in a vehicle

Abstract

The invention relates to a method for recognizing a construction site (10) in a vehicle (1), which has at least one microphone (2), the microphone (2) being an outward-directed microphone (2) by means of which acoustic signals (A) from an environment of the vehicle (1) are detected at least in a direction of travel (F) and are evaluated in a computing unit (3) in order to classify, with an ascertainable probability, at least one device (9) used on a construction site (10) as an object emitting at least part of the acoustic signals (A) and to recognize the construction site (10).

Description

[0001] Method for detecting a construction site in a vehicle

[0002] The invention relates to a method for detecting a construction site in a vehicle according to the preamble of claim 1 and to a vehicle according to the preamble of claim 10.

[0003] In the future, automated vehicles at SAE Levels 3 to 5 (e.g., trucks, cars, buses, etc.) will increasingly be on the roads. The vehicle will perform the driving task. Just like a human driver, the vehicle needs an understanding of its surroundings. Automated vehicles primarily use three sources of information for this:

[0004] On-board sensors that perceive the environment live,

[0005] - A-priori information recorded in advance and stored in an HD card (high-resolution card), and

[0006] V2X communication, which represents communication between the vehicle and, for example, other vehicles (V2V) or the infrastructure (V2I).

[0007] For various reasons, the focus is primarily on onboard sensors and HD maps. Both can provide an acceptable level of security. However, a problem arises when the HD map is no longer up-to-date.

[0008] Construction sites in particular pose a major challenge in this regard. On the one hand, the infrastructure here is often not as well developed as that of a normal road. On the other hand, construction sites are often not static but are in motion, as is the case, for example, with the application of a new road surface. Construction sites are sometimes set up overnight and their position and / or boundaries change. At the same time, driving conditions in construction sites are often worse than in normal traffic infrastructure. For example, lanes may be narrower, lane markings less clear, and guardrails and lane markings may be missing. A similar situation arises with short-term adjustments to traffic flow for road and road maintenance work (e.g., trimming plants and green spaces adjacent to the road), as such adjustments are set up and moved within a very short space of time.In such situations, automated vehicles must recognize that their map is no longer up-to-date and perform the driving task only based on the on-board sensors.

[0009] EP 2 879 92 B1 describes methods and systems for detecting a construction site using information from multiple sources. In one example, a computing device configured to control the vehicle may be configured to receive information from multiple sources related to detecting a construction site on the road on which the vehicle is traveling. Furthermore, the computing device may be configured to determine a probability of existence of the construction site on the road based on the information. Furthermore, the computing device may be configured to modify a control strategy associated with a driving behavior of the vehicle based on the probability; and to control the vehicle based on the modified control strategy.

[0010] The invention is based on the object of providing a novel method for detecting a construction site in a vehicle and a novel vehicle.

[0011] The object is achieved according to the invention by a method for detecting a construction site in a vehicle having the features of claim 1 and by a vehicle having the features of claim 10.

[0012] Advantageous embodiments of the invention are the subject of the subclaims.

[0013] A method for detecting a construction site in a vehicle is proposed, which has at least one microphone. According to the invention, the microphone is designed as an outward-facing microphone, with which acoustic signals from the surroundings of the vehicle are detected at least in one direction of travel and evaluated in a computing unit in order to classify at least one device used on a construction site as an object emitting at least some of the acoustic signals with a determinable probability and to detect the construction site, for example, using frequency and / or order analysis.

[0014] For example, bulldozers, excavators, jackhammers, rollers, hedge trimmers, tar machines, etc. can be classified as equipment based on typical properties of the acoustic signals.

[0015] In one embodiment, the vehicle is operated in an automated driving mode.

[0016] In one embodiment, a driving strategy and / or driving behavior of the vehicle is adapted based on the classified object and the probability, and a status of the detected construction site is determined.

[0017] In one embodiment, the driving behavior is changed so that the vehicle is operated in a degraded mode, for example at an adjusted speed.

[0018] In one embodiment, the construction site is classified and its status estimated based on the classification of the equipment.

[0019] In particular, the construction site can be classified according to its type, dynamics and / or activity.

[0020] The type of construction site can be, for example: night construction site, road clearance, bridge work, application of a new road surface, lane relocation, etc.

[0021] The dynamics of the construction site can, for example, be differentiated between static and dynamic, whereby a static construction site does not change in its boundaries and extent (e.g. bridge work) and a dynamic construction site changes (e.g. new road surface, keeping the road clear).

[0022] The activity of the construction site can, for example, make the distinction between active and inactive, whereby an active construction site is currently actively operated, so that it must be expected that people are present on the construction site and changes to the current database can occur at any time, whereas an inactive construction site is not currently in operation, for example, in the case of a construction site that is recorded in a digital map present in the vehicle or wirelessly available, no active device is identified based on the acoustic signal.

[0023] In addition, classification can be done using the time and usual working hours for the type of construction site.

[0024] For example, if the equipment used is classified as a hedge trimmer, then it can be assumed that the construction site is active and moving at a certain speed in the direction of travel of the vehicle.

[0025] In one embodiment, the vehicle forwards the detected information to a central database where it is made available to following vehicles.

[0026] In one embodiment, additional sensors, for example at least one camera, at least one radar and / or at least one lidar, can be used to classify the construction site.

[0027] In one embodiment, the additional sensors can be used to record an environment and update a map stored in the central database.

[0028] According to one aspect of the present invention, a vehicle, in particular an automated vehicle, for example a commercial vehicle, bus or passenger car, is proposed, which has at least one microphone and is configured to carry out the method described above.

[0029] The present invention enables the early detection of construction sites, as well as the detection of their active or inactive status via microphones. This information can be used to adjust driving behavior early on, increasing safety for everyone involved, especially for the vehicle, other road users, and construction site personnel.

[0030] Embodiments of the invention are explained in more detail below with reference to the drawings. In the drawings:

[0031] Fig. 1 is a schematic view of a method for detecting a construction site in a vehicle, and

[0032] Fig. 2 is a schematic view of the vehicle.

[0033] Figure 1 is a schematic view of a method for detecting a construction site 10 in an automated vehicle 1. Figure 2 is a schematic view of the vehicle 1. The automated vehicle 1 (for example a car, bus or truck) is in an automated driving mode (for example SAE Level 4 or 5) on a roadway, for example a public road, in ferry operation and has, among other things, at least one outward-facing microphone 2.

[0034] In a step S1, the automated vehicle 1 detects an acoustic signal A characteristic of a construction site 10 using the at least one microphone 2. In a step S2, the vehicle 1 evaluates the acoustic signal A in a computing unit 3. Machines and devices 9 used on construction sites 10 generate very specific acoustic signals A, which can be evaluated using frequency or order analysis and enable the classification of an emitting object with a determinable probability. This allows, for example, bulldozers, excavators, jackhammers, rollers, hedge trimmers, tar machines, etc. to be detected. The more acoustic signals A detected, the higher the confidence in the evaluation.

[0035] In a step S3, the vehicle 1 decides on an adjustment of the driving strategy and driving behavior (e.g., a speed) based on the results of the evaluation and the probability and determines a status of the detected construction site 10.

[0036] On the one hand, based on the early detection by means of the acoustic signals A, it is possible to detect that a construction site 10 is located on the route without visual perception. This allows the driving behavior of the vehicle 1 to be adjusted early on, and the vehicle 1 can be operated in a degraded mode (e.g., at an adjusted speed). If the construction site 10 is already present on a map, which is accessible, for example, via a wireless connection 8 to a central database 4, this is of course already known.

[0037] Based on the classification of the equipment 9 used, the construction site 10 can be classified and its status estimated. For example, if the equipment 9 used is classified as a hedge trimmer, then it can be assumed that the construction site 10 is moving at a certain speed in a direction of travel F of the vehicle 1.

[0038] In an optional step S4, vehicle 1 forwards the detected information to a central database 4. If the detected information, in particular about a moving construction site 10, is stored centrally, subsequent vehicles are provided with the information that their map will, with a certain probability, no longer be up-to-date in this area.

[0039] In support of microphones 2, further sensors, for example at least one camera 5, at least one radar, at least one lidar 6, etc., can be used to classify the construction site 10.

[0040] Relevant classifications for construction site 10 are mainly (more are conceivable): Type: Night construction site, road clearance, bridge work, new road surface, lane relocation,

[0041] Static: Construction site 10 does not move in its boundaries and extent (e.g. bridge works), Dynamic: Construction site 10 moves (e.g. new road surface, keeping the road clear),

[0042] - Active: Construction site 10 is currently in active operation. Therefore, there may be people on site, and changes to the current database may occur at any time.

[0043] Inactive: Construction site 10 is currently not in operation.

[0044] A construction site 10 can also be assigned a permissible combination of classes, such as active or dynamic. In this case, it can be assumed that the map is no longer up-to-date when a following vehicle passes through. Additionally, the classification can be based on the time of day and typical working hours for the type of construction site 10.

[0045] If the automated vehicle 1 is equipped with the sensors required to record map data, it can record the surroundings and update the map in a step S5.

[0046] The at least one microphone 2 can also be used, for example, to detect emergency vehicles. This allows security to be increased without increased costs.

[0047] In the case of an unknown construction site 10 which, on the basis of the evaluated signals, is considered impassable for the vehicle 1, for example a commercial vehicle, the vehicle 1 can be brought into a safe condition before the construction site 10 and will thus not lead to a blockage within the construction site 10, where stopping options are usually reduced or non-existent.

[0048] Not only the safety of the automated vehicle 1 currently passing through is increased, but also that of the following traffic.

[0049] List of reference symbols

[0050] 1 vehicle

[0051] 2 microphones

[0052] 3 Computing unit

[0053] 4 central database

[0054] 5 Camera

[0055] 6 Lidar

[0056] 8 wireless connection

[0057] 9 Device

[0058] 10 Construction site

[0059] A acoustic signal

[0060] F Direction of travel

[0061] S1 to S5 step

Claims

Patent claims Method for detecting a construction site (10) in a vehicle (1) having at least one microphone (2), characterized in that the microphone (2) is designed as an outward-directed microphone (2) with which acoustic signals (A) from an environment of the vehicle (1) are detected at least in one direction of travel (F) and evaluated in a computing unit (3) in order to classify at least one device (9) used on a construction site (10) as an object emitting at least some of the acoustic signals (A) with a determinable probability and to detect the construction site (10). Method according to claim 1, characterized in that the vehicle (1) is operated in an automated driving mode.Method according to claim 1 or 2, characterized in that, based on the classified object and the probability, a driving strategy and / or driving behavior of the vehicle (1) is adapted, and a status of the detected construction site (10) is determined. Method according to claim 3, characterized in that the driving behavior is changed such that the vehicle (1) is operated in a degraded mode. Method according to one of the preceding claims, characterized in that, based on the classification of the devices (9), the construction site (10) is classified, and its status is estimated. Method according to one of the preceding claims, characterized in that the vehicle (1) forwards the detected information to a central database (4), where it is made available to following vehicles. Method according to one of the preceding claims, characterized in that further sensors, for example at least one camera (5), at least one radar and / or at least one lidar (6), are used to classify the construction site (10). Method according to claim 7, characterized in that an environment is recorded by means of the further sensors and a map stored in the central database (4) is updated. Method according to one of claims 5 to 8, characterized in that the construction site (10) is classified with regard to the characteristics of type, dynamics and / or activity.Vehicle (1) comprising at least one microphone (2), characterized in that the vehicle (1) is configured to carry out the method according to one of the preceding claims.

Images