Technical Field

The present application relates to a method for a remote monitoring of a vehicle and to the corresponding stream adaptation entity configured for the remote monitoring of the vehicle. Furthermore, a computer program comprising program code and a carrier comprising the computer program is provided.

Background

Video Surveillance:

Video surveillance is commonly used in public spaces and generates large data volumes, most often at fixed locations but also in public transportation. The need for video surveillance in public transportation, on both rail and roads, for monitoring remotely passenger’s safety and behavior is expected to grow even larger as transportation is getting driver-less and remotely monitored and/or controlled.

Remote Driving:

A driver in a manually driven vehicle uses the field of vision offered by the front windshield along with that of the rear and sideview mirrors to maneuver the vehicle in various traffic situations. To provide a similar field of vision to remote operators, a vehicle that is remotely operable is expected to be equipped with multiple cameras and the resulting communication data rate increases with the number of cameras. Although a driver in a manually driven vehicle uses the field of vision offered by the windshield and the mirrors, at a given time the driver focuses on a small portion of the available field of vision.

Advanced Driver Assistance Systems:

These systems assist a human driver, who is fully responsible for the vehicle. In this context, the Day 1 C-ITS (Cooperative-Intelligent Transport Systems) services focus on increasing the human driver’s awareness of traffic environment by providing relevant information and warnings through V2X communications. Furthermore, the services can also provide some active assistance, for e.g., emergency breaking when the human driver does not respond to the warnings. Autonomous Driving:

In these systems, a vehicle is acting autonomously and is supervised/supported by the system. In this context, advanced C-ITS services may enable cooperation between autonomous vehicles (AVs) in addition to the functionality provided by the first day C-ITS services. The advanced services may range from intention sharing to cooperative path-planning and maneuvering..

Adaptive Streaming:

Adaptive streaming technique has been widely used for video streaming. Basically, the video is delivered chunk by chunk(e.g. ~4 second), each chunk of the video might have good or bad quality, e.g. resolution, according to the actual network/air interface congestion level. For instance, when the network/air interface congestion is low, the video chunk is encoded and delivered using high quality setup, when the congestion is high, the video chunk is encoded and delivered using low quality setup. The purpose of such adaptive streaming is to guarantee certain service continuity and deliver the video using limited network resources by best effort.

Methods for dynamic streaming rate adaptation have been proposed to adapt video rate to measured network performance. Self-clocked rate adaptation for multimedia (SCReAM) provides a rate adaptation algorithm for multiple streams through network congestion control and media rate control. To prioritize between different streams, streams can be assigned different scheduling weights.

Eye-tracking optimized video streaming:

A special case of adaptive video streaming is where the user field-of-view or even eye-gaze is used in the video stream adaptation, such that high quality is only delivered in the area where it matters to the viewer, i.e. around the viewers fixation points. In surrounding areas, the resolution can be decreased without any or only minor effects on user perceived quality. This way substantial bandwidth savings can be achieved. For instance, in [P. Lungaro et al, Gaze- aware streaming solutions for the next generation of mobile VR experiences. IEEE Transactions on Visualization and Computer Graphics, Vol. 24, No. 4, April 2018] a method for bandwidth optimization for 360 degree video delivery is presented. The method makes use of in-built eye-tracker in the user’s head-mounted display in combination with a codec configuration built on the High Efficiency Video Coding (HEVC) standard such that individual portions of video frames can be separately addressed and processed for near-instantaneous video quality adaptation depending on what part the user is looking at. Today, streaming of sensor data, which includes and is not restricted to audio and video data, does not consider the context, end-user interest, and the necessary features in the data. Consequently, the amount of data streamed from the source to the end-user or processing terminal is redundant and larger than necessary for the intended purposes. The number of sensors that are going to be deployed is going to increase dramatically in the future and if the current state-of-art methods for streaming the sensor data are continued to be used, the communication networks will be congested owing to the mismatch in available capacity.

Summary

Accordingly, a need exists to overcome the above-mentioned problems and to further improve the remote monitoring of a vehicle based on a plurality of data streams, such that the remote monitoring is also possible when the capacity may not be sufficient to transmit all sensor streams with the highest quality.

This need is met by the features of the independent claims. The dependent claims describe further embodiments.

According to a first aspect, a method for a remote monitoring of a vehicle based on a plurality of data streams is provided at a remote monitoring location. Each data stream is generated by a corresponding sensor provided in proximity of the vehicle and transmitted over a wireless network with a stream specific transmission rate to the remote monitoring location. The stream adaptation entity determines a predicted route of the vehicle and determines an event along the predicted route. Furthermore, it determines at least a first data stream of the plurality of data streams which is to be adapted based on the determined event. Furthermore, a priority with which the determined at least first data stream is transmitted over the wireless network is adapted based on the determined event.

With the knowledge of the predicted route and the event along the predicted route the plurality of data streams can be adapted such that one of the data streams, the at least first data stream focuses on the event along the predicted route so that at the remote monitoring location it can be made sure that the event along the predicted route can be precisely monitored.

Furthermore, the corresponding stream adaptation entity is provided comprising a memory and at least one processing unit wherein the memory contains instructions executable by the at least one processing unit. The stream adaptation entity is operative to determine a predicted route of the vehicle and to determine an event along the predicted route. Furthermore it is operative to determine at least a first data stream of the plurality of data streams which is to be adapted based on the determined event and to adapt a priority with which the determined at least first data stream is transmitted over the wireless network based on the determined event.

As an alternative a stream adaptation entity is provided configured for a remote monitoring of a vehicle at the remote monitoring location based on a plurality of data streams. The stream adaptation entity comprises a first module configured to determine a predicted route of the vehicle and a second module configured to determine an event along the predicted route. A third module is configured to determine at least a first data stream of the plurality of data streams which is to be adapted based on the determined event, wherein each of the data streams is generated by a corresponding sensor provided in proximity of the vehicle and is transmitted over a wireless network with a stream specific transmission rate to the remote monitoring location. A further module, a fourth module, is configured to adapt a priority with which the determined at least first data stream is transmitted over the wireless network based on the determined event.

Furthermore, a computer program comprising program code is provided to be executed by at least one processing unit of the stream adaptation entity, wherein an execution of the program code causes the at least one processing unit to execute a method as discussed above or as discussed in further detail below.

Additionally, a carrier comprising the computer program is provided wherein the carrier is one of an electronic signal, radio signal, and computer readable storage medium.

It is to be understood that the features mentioned above and features yet to be explained below can be used not only in the respective combinations indicated, but also in other combinations or in isolation without departing from the scope of the present invention. Features of the above- mentioned aspects and embodiments described below may be combined with each other in other embodiments unless explicitly mentioned otherwise.

Brief description of the drawings

The foregoing and additional features and effects of the application will become apparent from the following detailed description when read in conjunction with the accompanying drawings in which like reference numerals refer to like elements. Fig. 1 is a schematic architectural view of a system in which a vehicle is controlled at a remote monitoring location based on several data streams transmitted over a wireless network.

Fig. 2 is a more detailed view of the stream adaptation entity and its interaction with the remote monitoring location.

Fig. 3 is a schematic view of the vehicle traveling along a predicted route where an event occurs.

Fig. 4 shows an example flowchart of a method carried out at the stream adaptation entity in which the stream is adapted based on an event occurring along a predicted route.

Fig. 5 shows an example schematic representation of a stream adaptation entity configured to adapt a priority of a data stream based on a predicted route and an event along the route.

Fig. 6 shows another example schematic representation of the stream adaptation entity shown in Fig. 5.

Detailed Description of Drawings

In the following, embodiments of the invention will be described in detail with reference to the accompanying drawings. It is to be understood that the following description of embodiments is not to be taken in a limiting sense. The scope of the invention is not to be intended to be limited by the embodiments described hereinafter or by the drawings, which are to be illustrative only.

The drawings are to be regarded as being schematic representations, and elements illustrated in the drawings are not necessarily shown to scale. Rather the various elements are represented such that their function and general purpose becomes apparent to a person skilled in the art. Any connection or coupling between functional blocks, devices, components of physical or functional units shown in the drawings and described hereinafter may also be implemented by an indirect connection or coupling. A coupling between components may be established over a wired or wireless connection. Functional blocks may be implemented in hardware, software, firmware, or a combination thereof. Fig. 1 shows a schematic view of a system in which a user 80 at a remote location is monitoring a vehicle 60 based on data streams generated by different sensors 61, 62 and 63 in the neighborhood of the vehicle. The different data streams are transmitted over a wireless network 200, which in the embodiment shown can be implemented as a cellular network 222 with different cells. Each of the sensors 61 to 63 generates a real-time data stream which is transmitted over the wireless network 200 to a stream adaptation entity 100 provided at the remote location 50. The user or operator 80 has different display elements 71, 72, 73, and 74 with which the different data streams can be monitored. The vehicle 60 and the stream adaptation entity 100 may be each equipped with a transceiver, by way of example transceiver 40 shown at the vehicle 60. The transceiver 40 may comprise a mobile entity or user equipment, UE or any type of communication entity configured to transmit data streams over the wireless network to the remote monitoring location 50, especially the stream adaptation entity 100. The stream adaptation 100 is again able to communicate with the vehicle 60 or with the cellular network 200.

The wireless network may be any type of network allowing transmission of control messages and/or user data between entities connected to the network. The entities may subscribe to the network and each of the entities communicating over the network may be equipped with a subscriber identity module, SIM, which associates a user with a UE. The network, however, may also be a Wi-Fi or any other type of wireless network. Other short range technologies such as Bluetooth or ZigBee may also be used.

In the following, procedures for the dynamic adaptation of the data streams generated by the different sensors are discussed in more detail. The different data streams offered by the different sensors 61 to 63 are adjusted in a reactive and proactive fashion based on a combination of events detected at the information source close to the vehicle, by way of example the detection of a specific object. Furthermore, events detected at the user 80, also called operator, may be taken into account or information provided at any third party such as a traffic information available at a server.

In the following, it is assumed that the vehicle 60 is remotely monitored and possibly controlled by a human such as the operator 80 or by a machine operator. By way of example, the operator or user 80 may analyze the cameras provided on the vehicle or the operator may remotely steer a vehicle or supervise its autonomous driving. To this end, the multiple parallel real-time sensor data streams of the sensors 61 to 63 are transmitted to the remote location 50. The sensors 61 to 63 can include video cameras, radars, lidars, infrared sensors etc. The different sensors 61 to 63 may cover different areas and serve different purposes. By way of example, the cameras may have different fields of view, some of the cameras may be used for monitoring the environment outside the vehicle whereas other cameras may be used for the monitoring of the vehicle passengers.

Each of the different data streams can be transmitted over the network 200 with a stream specific transmission rate. This transmission rate can depend on the amount of data to be sent, by way of example can depend on the video resolution used or on the image frame rate. These multiple streams can pose a significant burden to the access network of the wireless network, especially in the uplink part which is typically less performant than the downlink. In case of a network congestion, the real-time sensor feeds may degrade in quality and even experience interruptions with obvious safety threats. However, in most situations, the sensor streams differ in relevance for the remote operator or user 80. Some of the streams may be of no relevance at all for a particular situation, wherein the same stream may be important and critical at another time in another situation.

By way of example, a video stream of a camera facing a bus exit door in a public transportation vehicle may not be of relevance when a detected traffic event on the road has to be avoided. However, it may be again of critical importance if the bus needs to be evacuated or if the bus approaches a bus stop where the leaving or entering of the people can be monitored at the bus exit door.

Further, it is possible that different parts within one sensor stream may differ in relevance. By way of example, only the part of a video stream that a remote human operator is looking at is useful and should be delivered at high resolution, whereas other parts are less relevant and may be delivered in a lower resolution. This inter alia motivates the optimization of the video feeds generated by the cameras to reduce the necessary data rates. The resolution may also depend on the fact whether a human operator or a machine operator is involved in the monitoring and controlling of the sensor streams. A machine-based object detection may require a lower video resolution compared to human operator or vice versa.

In the following, different options are explained which can be used to dynamically prioritize one or several of the video streams compared to other video streams. The adaptation of the video stream may be such that one specific video stream, by way of example a first video stream is amended such that the stream specific transmission rate with which the first video stream is transmitted to the remote location is increased compared to the other video streams. It may be increased such that it is higher than a defined threshold transmission rate. Furthermore, the transmission rate of other video streams, which are considered to be less relevant, may be decreased.

Furthermore, it is possible to determine the total available bandwidth that is available to transmit the different data streams to the remote location 50. This total bandwidth can then be distributed to the different video streams or data streams based on the relevance or priority. To this end, the different video streams may be categorized according to their importance, and the higher the importance is, the higher the bandwidth is that is used for the corresponding video stream from the total available bandwidth. In the same way, the lower the relevance is, the lower the corresponding bandwidth can be with which the corresponding data stream is transmitted to the remote location 50.

For the adaptation of the priority different pieces of information may be taken into account: a first piece of information can be information generated in the neighborhood of the vehicle describing an event occurring in the neighborhood of the vehicle. This first piece of information may consider information such as whether the vehicle is driving in the front or rear direction. This information can furthermore consider any other interesting, important or unexpected events occurring within or outside the vehicle in the near neighborhood. If the vehicle is moving forward, the sensor providing information about the direction and environment located in front of the vehicle may be transmitted with a higher resolution compared to a camera which is monitoring the rear part of the vehicle. If the vehicle is driving backwards, the sensor monitoring the rear part of the vehicle can be of higher importance so that the video stream monitoring the rear part gets a higher priority compared to the other video streams such as the video stream monitoring the front part of the vehicle. This piece of information is transmitted to the stream adaptation entity 100 and the stream adaptation entity can then adapt the priority of the different video streams accordingly and inform the vehicle how the different video streams should be transmitted via the wireless network. Furthermore, it is possible that in the environment of the vehicle a hazardous object such as a pedestrian is detected. As a consequence the relevant sensor data allowing a monitoring of this object are prioritized over the other video streams.

- A second piece of information that can be considered is information obtained by a network node including an application server which is aware of interesting/unexpected or potentially dangerous events along a predicted route. The vehicle can be equipped with a navigation system configured to monitor the position of the vehicle and configured to predict the route of the vehicle, either based on an input of the driver or based on routes often used by the vehicle. This second piece of information may also be generated at the vehicle itself, by way of example when the route is known by the vehicle. In one example, the vehicle will be crossing a hazardous zone, where a high accident rate is known or a school area. When the vehicle is approaching this zone the corresponding sensor streams may be adapted such that the priority of these streams is either increased or the bandwidth is such that a certain bandwidth is guaranteed for the stream which is able to monitor the hazardous zone. Another example of an event along the predicted route is the detection of the fact that the vehicle is traveling in a traffic jam and moving slowly. In such a situation, a high-resolution data stream may not be necessary as the vehicle speed may be very low or as the vehicle is even standing still. Furthermore, it can be detected that a basic car follow function such as an Advanced Driver Assistance System, ADAS, is operating in which the present vehicle 60 is automatically following the vehicle in front. Here, the required quality of the transmitted data stream is low and there is no need to provide high-quality and high-resolution data which require a large bandwidth. A further example is that the vehicle is waiting in front of a red traffic light. When the information is available how long the traffic light will stay red, the priority of the corresponding data streams monitoring the surrounding of the vehicle may be lowered for the predicted waiting time.

- Another piece of information is an operator-related information. When the operator or user 80 selects a certain data stream as important data stream, the corresponding priority of this data stream may be increased so that this data stream is transmitted with the required bandwidth. Here, the eye position of the operator may be tracked in order to see which part of the data stream the operator is currently focusing on. This part can then be transmitted in a higher resolution compared to the other parts or the other streams. Furthermore, machine-based object detection techniques may be used to detect objects in the provided data streams and the parts of the streams displaying the detected objects may be transmitted in a higher resolution than other parts of the same data stream or in the other data streams.

- A further piece of information is network-related information indicating the available or predicted network capacity. By way of example, when the network capacity is enough, all data streams may be transmitted with the same high quality, whereas in another situation when the network capacity is lower, some of the data streams may be prioritized over other streams. In the embodiment shown in Fig. 1 , the stream adaptation entity 100 is located close to the operator or user 80. However, it should be understood that this stream adaptation entity may also be implemented at the remote vehicle or in a controller network node such as in one node of the network 200 or may be provided in a distributed fashion in a cloud environment.

The different pieces of information can also be combined. By way of example, the parameters for the stream adaptations can be derived by combining the subjective intent or interest by the operator and the objectively important aspects of a service/use case in addition to the available network capacity.

By way of example, in a remote or a tele-operated driving the operator may be presented with several video streams originating from different sensors on the remotely driven vehicle. The gaze of the human driver can be used to assign different quality and priorities to the different video streams. However, this use case inherently requires that a remotely driven vehicle avoids collisions with other vehicles and objects. A collision prediction algorithm that is using the sensor data of the vehicles or which is analyzing the video streams at the vehicle or the human operator can also be used to control the parameters for the stream adaptation to increase the quality of the critical video stream to bring the attention of the human operator to a potential collision.

The decision how to prioritize the different data streams can exploit the different pieces of information given above and, in case of a potential network congestion or other network traffic reduction constraints such as connectivity costs, can dynamically adapt the rate at which at least some of the parallel transmitted sensor data streams are prioritized based on the decision how to prioritize each of the streams. The prioritization decision is based on at least some of the pieces of information mentioned above. Furthermore, the prioritization depends on the degree of traffic reduction constraints. This can mean that at no data traffic congestion all sensor streams may be transmitted at full bandwidth or all streams having a certain relevance above a threshold may be prioritized whilst at a congestion only the most critical stream or the most critical streams are prioritized.

The stream adaptation entity may obtain site information regarding the traffic event ahead on the route and prioritize forward facing sensors over the inward or rearward cameras as mentioned above. For the prioritized forward facing cameras the video stream is optimized so that a high resolution is only delivered in the eye gaze position of the remote operator. The outcome of the method discussed above is that the stream adaptation entity indicates explicitly the data from which sensors should be sent using what data rate and/or compression scheme. In case the stream adaptation entity is implemented in the network 200 the rate adaptation command can be sent from the network to the vehicle, e.g. to the UE implemented in the vehicle. It is possible that some data streams from some sensors might not be requested by the stream adaptation entity at all.

The real time sensor data is sent to the network and the stream adaptation entity specifies different qualities or data rates according to the adapted priority. Sometimes it is still needed to store the raw data at the stream adaptation entity due to legislation issues, by way of example for the liability in case of a vehicle accident. For this purpose the original high-quality raw data can be sent to the stream adaptation entity in a lower than a best effort fashion, where the data is transmitted only when the available capacity exceeds the current need, by way of example as background data. To avoid the duplication the stream adaptation entity or the UE can check which sensor data has not been sent or has been sent after compression or down sampling. The vehicle will then send those data in high quality or only the compensating part to the stream adaptation entity or the remote location. At the remote location the different data can then be merged with the already received real-time data so that a complete high-quality sensor data record is obtained.

Fig. 2 shows a more detailed view of the exchange of information between the stream adaptation entity 100 and the vehicle 60. The stream adaptation entity receives information from network 200 about the network load and the available capacity. Further, more information of different services is collected, by way of example information about the different bus stops, information about traffic congestions, information about red light phases etc. In addition, information from the operator or user and the human-machine interface may be considered and all these pieces of information are used to determine how the different data streams are prioritized, which data stream is transmitted with the full bandwidth, which data stream is not transmitted at all, which data stream is transmitted with reduced bandwidth and/or which data stream is prioritized over other data streams.

When it has been decided, which data stream is transmitted with which bandwidth and which compression scheme, the corresponding information is transmitted to the vehicle and the corresponding information is used at the vehicle for controlling the sensors and the data streams generated by the sensors. The sensors 61 to 63 either generate the sensor data streams at the determined bandwidth and/or resolution, or the sensors provide the predefined data quality and it is then decided whether the sensor data is transmitted at the initial stream quality or whether the bandwidth is reduced based on the determined priority. The decision how to adapt the different sensor streams is further more influenced by the local context detection such as the information detected directly in the vehicle or around the vehicle. When it has been determined how the different data streams are transmitted from the vehicle to the stream adaptation entity 100, the corresponding data streams are transmitted to the remote operator or to a corresponding server located at the remote operator.

Fig. 3 shows an example schematic view of the vehicle 60 traveling along the predefined route 30. Further, an additional part of the route 31 is shown, however, the vehicle currently assumes that the vehicle is traveling along route 30. Furthermore, the information is available that an event 33 is occurring along the predicted route. This event can be a traffic jam, an accident, or any other hazardous situation. Based on this information the stream adaptation entity can adapt the different sensors such that the sensor which can best monitor the event 33 is controlled in such a way that the stream specific transmission rate is high enough that the event 33 can be evaluated correctly.

Fig. 4 summarizes some of the steps carried out at the stream adaptation entity in the above- discussed method. According to a first step S91, the predicted route such as route 30 is determined. In step S92, the event such as event 33 is detected furthermore, in step S93 at least one of the data streams of the several data streams is determined which is to be adapted based on the determined event. By way of example, it can be the data stream of the sensor which can fully monitor the detected event 33. Step S94 then comprises the adaptation of the priority of the data stream determined in step S93. This can mean that either the priority of the data stream which can correctly monitor the event at the route is increased and/or it can mean that the priority of the other data streams which are less relevant for the monitoring of the event is decreased. Furthermore, both of these measures can be taken.

Fig. 5 shows a schematic architectural view of the stream adaptation entity which can carry out the above-discussed adaptation of the priority of the different data streams. The entity 100 comprises an interface 110 which is provided for transmitting the user data or control messages to other entities, by way of example the information how the different data streams should be transmitted from the vehicle to the stream adaptation entity. The interface 110 is furthermore configured to receive inter alia the different data streams at the determined bandwidth and/or compression scheme. The entity 100 furthermore comprises a processing unit 120 which is responsible for the operation of the entity 100. The processing unit 120 comprises one or more processors and can carry out instructions stored on a memory 130, wherein the memory may include a read-only memory, a random access memory, a mass storage, a hard disk or the like. The memory 130 can furthermore include suitable program code to be executed by the processing unit 120 so as to implement the above-described functionalities in which entity 100 is involved.

Fig. 6 shows another architectural view of a further embodiment of a stream adaptation entity. The stream adaptation entity 300 comprises a module 310 configured to determine the predicted route of the vehicle. The route may be determined internally within the vehicle or may be determined remote from the vehicle based on other factors such as a predefined schedule route, e.g. for a public transportation vehicle. A module 320 is configured to determine an event along the predicted route. This event can contain a traffic jam, an accident, or a stop of a public transportation vehicle at a defined stop etc.

A module 330 is provided configured to determine at least one of the data streams, named first data stream which has to be adapted based on the determined event. Furthermore, a module 340 is provided configured to adapt the priority of at least the first data stream.

From the above said some general conclusions can be drawn:

(here we summarize the dependent claims)

The adaptation of the priority of the data stream can mean that the corresponding stream specific transmission rate is adapted with which at least the determined data stream is transmitted to the remote monitoring location such that the first stream specific transmission rate is increased.

The increase can be such that it is at least higher than a defined threshold transmission rate.

The adaptation of the priority can furthermore mean that the stream specific transmission rate of the determined data stream with which at least this data stream is transmitted to the remote monitoring location is adapted such that a total available bandwidth is determined available to transmit the plurality of data streams, wherein a defined part of this total available bandwidth is reserved for the data stream for which it is determined that it has to be adapted based on the determined event.

The adaptation can furthermore mean that a stream specific transmission rate is reduced for at least one of the remaining of the plurality of data streams relative to this first data stream which is determined based on the event along the route. By way of example, as discussed above if it is determined that the vehicle is driving in the rearward direction, it may be possible to reduce the stream specific transmission rate for the data streams of the sensors which provide information in the forward direction of the vehicle.

Furthermore, it is possible to determine operator- related information from the remote monitoring location describing an event occurring at the remote monitoring location. The priority of the at least first data stream can then be adapted based on the determined operator- related information.

The operator- related information can mean that the operator is focusing on one of the data streams and wherein this data stream is selected as the first data stream and the priority of this first data stream, on which the operator is focusing on, is increased relative to the remaining of the plurality of data streams.

Furthermore, it is possible that the transmission capacity of the wireless network is determined wherein the priority of the determined at least first data stream is adapted only when the transmission capacity is lower than a threshold. This can mean, when the transmission capacity is high enough, all the data streams are transmitted with a corresponding default transmission rate which can depend on the sensor characteristics and its resolution.

Furthermore, it is possible to determine neighborhood information describing an event occurring in the neighborhood of the vehicle. One of the data streams is then selected as the at least first data stream with which the monitoring of the event occurring in the neighborhood of the vehicle is possible.

The event occurring along the predicted route can comprise a dangerous or hazardous situation for the vehicle occurring in proximity of the vehicle or along the predicted route. Said one data stream is then selected as the at least first data stream with which the monitoring of the dangerous situation is possible.

The neighborhood information or the event along the predicted route can comprise a traffic congestion. When it is detected that the velocity is slower than a threshold, the priority of the at least first data stream can be adapted such that a transmission rate of the plurality of data streams including the first data stream is lowered.

Furthermore, it can be determined if the vehicle is operating in a vehicle following operating mode in which the vehicle is automatically following the direct vehicle in front. If this operating mode is detected, the priority of the at least first data stream may be adapted such that a transmission rate of the data streams is lowered. As discussed above, in this vehicle following operating mode, it may be only necessary to correctly identify the position of the vehicle directly in front. Other sensory data may not be needed in a high resolution.

The neighborhood information can furthermore comprise determining that the vehicle is not moving due to a red light, where, in a waiting period, it is determined how long the red light will stay red. The priority of the determined at least first data stream can then be adapted such that a transmission rate of the data streams or of most of the data streams is reduced for the determined waiting period.

Furthermore, it is possible that a priority level for each of the plurality of data streams is determined wherein the stream specific transmission rate is reduced for data streams having a priority level below a threshold.

Furthermore, it is possible to determine a priority level for each of the plurality of data streams, wherein the stream specific transmission rate is adapted taking the priority levels of the plurality of data streams into account.

Furthermore, it is possible to determine a vehicle specific information describing an event occurring within the vehicle wherein the priority of the determined at least first data stream is adapted taking into account the vehicle specific information.

The vehicle specific information can comprise determining a driving direction of the vehicle, wherein one of the plurality of data streams is determined providing information in the direction of the driving direction. Said one data stream is then selected as the first data stream and the priority of this first data stream is increased relative to the remaining of the plurality of data streams. This can mean that the data stream monitoring the driving direction is prioritized over the other data streams not monitoring the driving direction.

For at least some of the plurality of data streams the stream specific data rates and compression schemes are determined that are used to compress the corresponding data streams wherein the determined stream specific data rates and compression schemes are signaled to the vehicle.

The vehicle can be a vehicle stopping at predefined stops along the predicted route and can comprise at least one image sensor providing an image data stream of a door of the vehicle used to enter or exit the vehicle. This image data stream is then the at least first data stream and the priority of this at least first data stream is adapted such that the priority is increased when the vehicle approaches each of the predefined stops and is lowered after the vehicle has left each of these predefined stops.

Furthermore, when it is determined that the priority of the at least first data stream has to be increased together with the priority of at least one further data stream from the plurality of data streams, a total available bandwidth available to transmit the plurality of data streams is determined and a relevance of the at least first data stream and the at least one further data stream is determined wherein the priority of the at least one first data stream and of the at least one further data stream is adapted based on the total available bandwidth and the corresponding relevance.

Furthermore, it is possible that at least a first data stream from the plurality of data streams is received in at least two different stream parts wherein the two stream parts have different stream specific transmission rates. The two different stream parts can then be merged to a merged stream part and the merged stream part may be stored at the stream adaptation entity. As discussed above, it may be required to store the data streams at the stream adaptation entity for legal reasons. Accordingly, the stream adaptation entity then receives two different stream parts, wherein the higher-quality stream part may be transmitted not in a real-time transmission but with a best-effort high quality in the background. The stream adaptation entity can then merge the already received lower-quality part with the later received higher-quality part.

Summarizing, the discussion above helps to reduce the network load for specific applications while still assuming that the information at the remote location is available that is needed to fully monitor the vehicle from the remote location.