The MEC-View system approach mainly bases on the evaluation of a secondary road side sensor setup. The road side sensors expand the field of view of the on-board sensors of the automated vehicles in order to gain a complete local environment model including all relevant objects.

In complex urban traffic scenes some of these relevant objects could be occluded for the on-board surround sensors. In this case, the MEC-View road side sensor setup may serve as a virtual mirror to amend the local environment model .

The MEC-View system architecture also includes a high-precision digital map, a high-performance mobile network, and a Mobile Edge Computing (MEC) server to allow for a reliable data fusion of objects derived form both, the road side and the on-board environment models.

The system concept will be implemented at a test site in the city of Ulm, in order to prove the MEC-View use case: automatically entering a priority road at an urban intersection on real traffic conditions.


Arch en V1



Automated driving in complex urban environments is limited due to occlusions of relevant road users or obstacles – in these situations the performance of vehicle bound surround sensor systems is restricted as a matter of principle, which either cannot be compensated by car-2-car connectivity in scenarios of incomplete sensing capability and/or connectivity of the overall vehicle fleet.


To tackle this problem, the public funded project MEC-View focusses on the evaluation of a complementary road side sensor system and a high-precision digital map of the driving environment in addition to the sensor systems and processing capability of an automated vehicle. Based on the road side sensor objects a mobile edge computing (MEC) server frontend delivers a local environment model via a prototype LTE-G5 mobile network to the automated vehicle.


The MEC-View project strives for a save and an efficient automated driving in complex and challenging urban situations. Moreover, the system provides an improved perception of vulnerable road users, e.g. pedestrians, cyclists and motor bikers.


logo light AP1 - MEC-View System
logo light AP2 - Digital Map

AP1 small v2

  • Requirements engineering
  • Project management
  • Pilot installation

AP2 small

  • Precise digital map
  • Dynamic track clear detection
  • Traffic analysis and simulation
logo light AP3 - Communication Network
logo light AP4 - Infrastructure Sensor System

AP3 small

  • Mobile radio network
  • MEC-Server HW
  • Communication units

AP4 small

  • Road side sensor
  • Setup and integration of street lights
  • Ground-truth sensor system
logo light AP5 - MEC-Server  logo light AP6 - Automated Vehicles

 AP5 small

  • MEC-Server SW architecture
  • SW implementation
  • Deployment and test

 AP6 small

  • Vehicle integration
  • Automated driving functions
  • Use case demonstration

At present, the automotive OEMs and suppliers prepare the automated driving to provide a more efficient and safe road traffic. The automated vehicle takes over control like a "Chauffeur" and the driver turns into a passenger. The vehicle accelerates, brakes, applies the turn indicators, and changes the lanes without any action needed by the driver.

On highways automated driving has been successfully validated now for some years. Prototype vehicles take over control and drive automatically and independently. The market launch of the "highway-pilot" is expected early at the next decade.

However, the highway situation without oncoming vehicles nor crossing or turning traffic cannot easily be transferred to the urban environment. In complex urban traffic scenarios different road users and a variety of unexpected conduct must be considered: heavy traffic and closely passing vehicles, children lingering on cross-walks, cycle messengers suddenly changing the lane, urban busses leaving bus stops on the spur of the moment - just to name a few.

Additionally, on-board sensors of an automated vehicle can neither detect all of the road useres nor driving areas in typical urban situations due to occlusions by obstructions or barriers like heading or parking cars, or any existing development at the way side.Therefore, the automated vehicles need additional information about the environment from other sources than their on-board surround sensor systems.

At a dedicated test site in the City of Ulm the MEC-View project partners investigate on what automated vehicles need in order to drive safely in a mixed traffic of conventional vehicles and other road users in an urban environment.

Today's automated vehicles are equipped with a variety of diverse surround sensor systems by means of RADAR, Video, ultrasonic, and LiDAR technologies, in order to gain a full 360-degree surround view. However, these systems are not able to detect a cyclist behind a truck at a crossing nor a pedestrian approaching a cross-walk from behind a wall.

In future, communal development authorities will install video or other sensors at street lights and sites which are important for the improvement of the local traffic situations. The project MEC-View uses the data of these road side sensors to make automated driving accessible in urban environments.

The MEC-View partners develop software and hardware to process the video frames and sensor data coming from the traffic lights. Extracted object features, e.g. positions and velocities of the road users, are associated with high-precision digital maps and are transferred to the automated vehicle in real time through a mobile radio network.

The data from the road side sensors is merged with the on-board sensor data of the automated vehicle in order to get an accurate and complete model of the surrounding environment including all relevant road users, obstacles, and traversable area. The automated vehicles are now able to master the challenging situations in complex urban traffic, e.g. entering a priority road or passing a rotary intersection.

In order to meet the positioning requirements of the road users in a local environment model, the sensor data need to be transferred in real-time operation. The MEC-View team uses a low-latency prototype LTE/5G mobile radio network and a mobile edge computing (MEC)-server to satisfy the timing demands. The MEC-server hosts an algorithm for the data fusion and tracking of the road side sensor data and associates these data on a precise digital map. On the basis of the LTE/5G network, the server provides the local environment information to the automated vehicle as quick as a flash.

In the future, traffic control and information systems of urban administrations could be equipped with MEC-servers to provide this local environment model information. Precondition for the automated driving on the MEC-View approach is the area-wide coverage of the LTE/5G mobile radio network.