NIST Vehicle Teleoperation Forum

Virtual Conference
November 20, 2020

NIST Vehicle Teleoperation Forum

On Friday November 13, 2020, the National Institute of Standards and Technology (NIST) will hold a full day virtual forum designed to bring leaders from the industry, academia, and government agencies to discuss the challenges, opportunities, and potential paths forward in vehicle teleoperation. The forum presenters will deliver their views on the future of automated driving and vehicle teleoperation, the roles of cloud computing and 5G in vehicle teleoperation, and the key challenges to be addressed.

The goal of the forum is to identify the key issues in vehicle teleoperation that can be best addressed by collaboration. Examples of these areas include performance and safety measurements, architecture, and standardization needs.

The goal of producing a fully automated vehicle (AV) represents a culmination of automotive technology with sophisticated control software, AI-based models, ultra-low latency and reliable communications, and operational management systems occurring on the vehicle, on the cloud and over the network using a standards-based approach. A crucial part of the AV ecosystem is vehicle teleoperation (remote driving) and accordingly is receiving growing industry attention. Teleoperation is playing an important role in helping self-driving cars navigate through unusual and difficult situations and providing emergency assistance to them.

Incumbent leaders and many startups in automated driving, along with ride-sharing companies, goods delivery companies, and cloud providers have been developing and advancing teleoperation systems. Current teleoperation systems use human teleoperators to guide or drive vehicles through difficult situations. Teleoperated vehicles can also provide new services such as teleoperated taxis and teleoperated delivery services. Teleoperation will also help other actors, such as law enforcement officers, automotive technicians, and parking attendants, interact with automated vehicles.

Network service is a crucial part of vehicle teleoperation and other advanced automated driving functions such as cooperative driving. Supporting remote driving is an important goal of 5G systems that are being commercially deployed worldwide. Telecommunications companies and the car industry currently test vehicle teleoperation over 5G networks.

Over time, as artificial intelligence (AI) is introduced into vehicle teleoperation, the teleoperation system will become increasingly more capable. The AI systems will provide advanced assistance to human teleoperators, for example, by predicting network conditions and providing guidance. As the AI develops and interacts with the human operator, cognitive scientific approaches can be applied to develop new interfaces and approaches for the human/AI system. These future AI systems in the cloud, enabled by advanced networking technologies such as 5G, may also provide automated teleoperation services.

Vehicle teleoperation can benefit from voluntary consensus-based standards, which have been a crucial part of the technology landscape, to ensure trust and interoperability.

To realize the full potential of vehicle teleoperation, many issues need to be addressed. For example:
* What roles will teleoperation play in the future in automated driving?
* What will be the key challenges in vehicle teleoperation? For example, how to measure and validate the performance and safety of vehicle teleoperation? How to ensure safe vehicle teleoperation under poor network conditions? What unique security and privacy challenges may arise in vehicle teleoperation?
* What will be the key enabling technologies for supporting human teleoperators today and enabling AI-powered automated teleoperation in the future? For example, AI? cloud computing? 5G? Human cognitive models?
* Which aspects of vehicle teleoperation can benefit from standardization? For example, taxonomy? performance and safety metrics? networking requirements? interfaces with teleoperator, vehicle, and passenger?