COMMENT: Safe teleoperation is built on a myriad of technologies

Is teleoperation the route towards autonomous driving? Amit Rosezweig explores the key challenges to this approach

Teleoperation, which enables remote vehicular operation via public or private networks, is becoming the de-facto norm to enable the deployment of autonomous vehicle technology. With real time situational awareness, the human teleoperator can safely make the right decisions. Sounds simple? Not exactly.

Teleoperation relies on public cellular networks (LTE, 5G) and/or dedicated WiFi networks to transfer information between the vehicle and the teleoperation centre, with minimal delay. The key to successful operation is continuous two-way data streaming, regardless of shifting network conditions. Herein lies the challenge: neither LTE nor WiFi are designed to support such high bandwidth and ultra low latency communication from a moving vehicle, across a vast geography. Meanwhile, 5G will take years to roll out in entire cities and countries, just like with 4G. But there are immediately relevant use-cases for teleoperation.

The key to successful operation is continuous two-way data streaming, regardless of shifting network conditions

Teleoperation overcomes network limitations by offering sophisticated bonding of multiple networks, network-aware video transport, and efficient compression technology. These technologies work in tandem to optimize LTE and/or WiFi performance. Making optimal network bonding and video transport possible requires a blend of five discrete technologies:

  1. Modem bonding: Several modems working in unison provide improved throughput, redundancy and network coverage by minimising packet loss and latency. Using a packet scheduler, the channels are optimised to route packets via the most efficient channels in real-time.
  2. Dynamic Forward Error Correction (FEC): Lost packets are reconstructed at the receiving end without re-transmission of data. This minimises latency while enhancing the reliability of every channel. The dynamic part of the FEC algorithm minimises the overhead throughput needed for packet reconstruction.
  3. Dynamic receive buffer: Quality video transmission normally requires a buffer greater than 300ms. Based on network conditions, a dynamic receive buffer can reduce buffer time by up to 80%. When network conditions are good, buffer size can be as small as 5ms, and when conditions are suboptimal, it can max out at 50ms.
  4. Real-time monitoring: Dynamic tools require reliable real-time monitoring of changing network conditions. The system must know the received power in each antenna, network latency, packet loss, throughput and more. This data collector refreshes many times per second to keep up with network variance, and the system can respond immediately to these measurements.
  5. Adaptive video compression: Multiple raw video feeds are ingested, compressed and sent to the teleoperation centre. With real-time network input from the data collection module, video stream quality is dynamically adapted. For example, if the upload bandwidth (speed) drops below 4Mbps, video resolution is purposely degraded to maintain an always-on video stream. Adaptation requires no resetting of the video stream and happens within a single video frame.

Network bonding and video compression are just the tip of the iceberg. A reliable teleoperation platform also requires embedded cyber security, safety features, methods to provide remote assistance, situational awareness technology, and a well-designed user experience (UX).


The opinions expressed here are those of the author and do not necessarily reflect the positions of Automotive World Ltd.

Amit Rosezweig is Chief Executive of teleoperation software company Ottopia

The Automotive World Comment column is open to automotive industry decision makers and influencers. If you would like to contribute a Comment article, please contact editorial@automotiveworld.com