This project investigates the potential of remote-controlled vehicles as a transition to fully autonomous driving.

Background and motivation

 

Since the 80’s, research on the subject of “Tele-Operated Vehicles” has been increasing significantly. In addition to the early successes of the EU-Project PROMOTHEUS from the 90’s, the competitions of the U.S. Department of Defense (Darpa Challenges) have left a mark on the public perception. After the Institute of Automotive Technology devoted itself as part of a research topic from the DFG-Collaborative Research Center, it can be stated as a summary that autonomous driving is possible, but only in certain and less complex situations. Some highway or congestion- or Stop-and-Go-traffic scenarios partly fulfill these conditions. For all other scenarios, and especially in the field of urban traffic, it can be definitely said, that in the next few decades no autonomous vehicle will be driving on public roads. The reason is that complexity is too large, including many different road users and confusing road topology. Furthermore, the legal aspect forbids any autonomous vehicles.

Tele-operated road vehicles

 

Because of the reasons mentioned above, future research at the Institute of Automotive Technology is focused on “Tele-operated Vehicles” and “Semi-Autonomous Vehicles”. This approach, borrowed from Robotics, consists of bringing a human driver into the control loop in order to deal with the complexity of the road. This approach is already been used in the deep sea and space exploration. “Tele-Operated Driving” means in this context that an external operator drives the vehicle using a live-stream video. Here, the operator is connected to the vehicle via wireless data transmission, such as WLAN or cellular network (UMTS, LTE). A server handles the communication and can in addition perform computationally-intensive tasks or provide additional information from the internet. Further, a Car2Server-Communication is simultaneously developed, which apart from the tele-operation, is able to provide information to the driver, like road situations lying ahead.

Approach

 Applications for tele-operated road vehicles could be all forms of vehicle provision, like Car-Rental or Car-Sharing projects. Thereby, the vehicle could be delivered to the client to the desired location through tele-operation without having a staff to physically travel this route. This approach could considerably accelerate the spread of electric vehicles in urban areas through Car-Sharing concepts. In order to investigate the feasibility of “Tele-operated Driving”, an experimental vehicle is being built at the Institute of Automotive Technology. The “Tele-operated Road Vehicles” are primarily influenced by three issues to be investigated in the future:

1. The data transfer of both video data and control signals will lead to a time delay. The question is: How big is the permissible time delay for a safe driving and how can it be controlled?

2. Since the operator is not present in the vehcle, the question about what additional information does the operator need in order to drive a vehicle just as well as an on-board driver needs to be asked.

3. "Tele-operated Driving" is based on a wireless data transmission between the vehicle and the server. In doing so, an error-free data transmission cannot be guaranteed, which raises the question of an appropriate safety concept.

All these subjects are to be investigated and discussed here at the Institute of Automotive Technology.

If you are interested in the field of "Tele-operated Driving" and ethusiastic about new mobility concepts, you can get in touch with us directly under georg@ftm.mw.tum.de or feiler@ftm.mw.tum.de.

Teleoperated driving videos:

In August 2013 we showed at a local press event the possibilites of teleoperated driving in combination with our concept electric car visio.m:

Youtube:

Süddeutsche Newspaper: