Teleoperation of autonomous vehicles is the ability to remotely drive or assist a
self-driving car. Most leading companies in the industry believe that to bridge the gap between current self-driving capabilities and the requirements needed for widespread adoption of autonomous vehicles, there is a need to have teleoperation capabilities for assisting
self-driving cars, in situations of ‘edge cases’ – where the
autonomous vehicle software stack has low confidence level in its ability to perform the correct action, or when the vehicle needs to operate outside of its standard operating parameters. Without remote assistance, in such situations the self-driving car would transition to a minimum risk maneuverer (MRM) which is usually to stop. Many AV companies plan on using teleoperations as part of their rollout for self driving cars. Examples of companies that have stated they will deploy, or currently deploying teleoperations solutions include Voyage.auto,
Denso,
Waymo,
GM Cruise,
Aptiv,
Zoox. Teleoperation of autonomous vehicles includes privately owned self driving car use cases, such as self parking assistants, shared mobility use cases, e.g. in robotaxis and autonomous shuttles and industrial use cases, for example autonomous forklifts. There are two main modes for teleoperation of autonomous vehicles: Remote driving, also called "direct driving" – where the remote operator performs the dynamic drive task, i.e. drives the car remotely, controlling the car's
steering,
acceleration and
braking systems. Remote assistance, also called "high level commands" – remote operators supervise the vehicle, and provide instructions, approve or correct the vehicle path, without actually performing the dynamic driving task. Some companies deploy a combination of both concepts, depending on the use case. Examples of companies that provide solutions in the field of teleoperations are DriveU.auto, Roboauto, Pylot (now part of Fernride), Ottopia, Designated Driver, Soliton Systems, and Transitive Robotics. ==See also==