NTT’s autonomous vehicle development has taken a new turn, with the introduction of a working prototype technology designed to enable advanced remote driving.
The company formed NTT Mobility Inc. in November of this year to advance the autonomous driving business, leveraging skills in autonomous vehicles, maintenance, remote monitoring systems, and on-site response. Technologies in this space operate what is known as Level 4 deployment, i.e., a vehicle with high automation that can drive entirely (within a specific geographic area, or under certain conditions) without any need for human intervention.
But this is not Waymo, this is something different.
Driving Around Mount Fuji
Demonstrated at the company’s NTT R&D Forum showcase symposium this month, advanced remote driving enables a user to sit in a video game-like Recaro bucket seat at the wheel behind three connected widescreen monitors. An automatic drive engine can then be controlled via an accelerator and brake pedal with the real-world vehicle (located in this instance some 80 miles away on a test track at the base of Mount Fuji) moving its way around the road surface.
A safety driver was located in the vehicle at all times (possibly the worst job in the world?), just in case users experiencing this still-nascent technology decide to hit the [electric] gas too hard or perform a screwy steering manoeuvre.
Low-Latency Video Transmission
In an environment where screen buffering could be a matter of life and death, NTT’s own Innovative Optical Wireless Network (IOWN) is responsible for providing the low-latency video transmission connection needed to keep the vehicle on the straight and narrow.
NTT says its goal is to achieve stable remote driving through proactive communication control and efficient video transmission.
“In remote driving, it is necessary to transmit real-time video of the vehicle’s movement to a remote location. Operations can be hindered by delays or interruptions in the video, depending on the communication quality. Stable communication technology is enabled here using ‘aggregation control’ for multiple access according to wireless quality prediction,” noted NTT, in a technical briefing document.
High-Quality Compressed Data
Other software-based mechanics at work here include stable video transmission technology that transmits data highly in a compressed format. A communications stabilization service is also crucial for maintaining connectivity in environments where wireless quality fluctuates (e.g., due to base station switching or area interference).
“[This solution] uses multipath connections across multiple communication lines between the vehicle and the remote monitoring system. It proactively predicts changes in wireless quality using AI/machine learning (using IOWN technology like Cradio), and then controls the multipath connections to ensure high connectivity and reduce the risk of video interruptions,” specifies NTT.
NOTE: NTT IOWN Cradio is a multi-radio proactive control technology that understands, predicts and controls a wireless communications environment to provide an optimal user connection.
Parametric Object-Recognition-Ratio-Estimation
NTT has also developed the parametric object-recognition-ratio-estimation model. This technology automatically evaluates whether the video transmitted from autonomous vehicles is clear enough for human operators to safely detect obstacles in real-time, enhancing overall autonomous driving safety.
