Development of SAE Level 4 and Level 5 autonomous driving has been slower than expected (in Level 4, the vehicle can drive itself under limited conditions and in Level 5, it can drive itself under all conditions). Why? The commercial emergence of self-driving cars has been delayed due to 1) technical challenges in handling unpredictable scenarios, 2) the need for near- perfect safety reliability, 3) unpredictable edge cases such as bad weather and 4) complex regulatory hurdles. SAE Level 4 autonomous vehicles are operating commercially in a few U.S. cities (for example: Phoenix, San Francisco, Los Angeles and Austin) via robotaxi services. Level 5 remains theoretical, with widespread adoption believed to still be 5-10 years away.
In the meantime, to allay public concerns about the technology and the absence of human drivers behind the wheel, teledriving is becoming a viable alternative. Teledriving allows a human operator to drive a vehicle remotely from a distance using cameras, sensors and 5G/4G connectivity. It enables services like remote-delivered car rentals where a human can remotely drive the car to the user without being inside.
The teledriver setup at a remote center includes a steering wheel, brake and accelerator pedals accompanied by monitors that display the car's surroundings in near real-time (a delay of even a few hundred milliseconds could prove fatal, given the time it would take for the driver to pull out safely into traffic or to react to an emergency). The teleoperator gets the view from the car as if they are looking through the windscreen, mirrors and side windows. The data feed includes sounds so teledrivers can hear approaching emergency vehicles.
Delivering Rental Cars
As noted, teledriving can be used for bringing rental cars to customers. Instead of a customer walking up to a car rental parking lot, an EV is remotely driven to a customer's location once it's requested via an app, allowing renters to jump in and use the vehicle as a regular rental car. Remote drivers operate in a controlled, professional environment free from distractions, and their driving behavior is continuously monitored. Vehicle drop-off occurs in a similar fashion: once a customer is done with the EV, they can exit and let the remote driver take over to park or pick up another customer.
Vay, a German mobility company, has been operating its remotely driven car rental service in Las Vegas since January 2024. Vay uses remote drivers to deliver EVs to customers. It has opened a bigger local office in downtown Las Vegas, grown its fleet to 175 vehicles on the road and has surpassed 60,000 rides. Vay allows customers to request an EV to be remotely delivered to their location. When it arrives, the remote driver disconnects from the vehicle and the user takes over, driving it like a regular car. At the end of the trip, they exit the vehicle, and a remote driver resumes control, eliminating the time-consuming search for parking.
The larger Vay fleet improves availability, reduces average wait times to less than 10 minutes for vehicle delivery and enables broader coverage across the city as customer usage continues to increase. The company has extended its remotely driven delivery service into Northeast Las Vegas, adding to the areas already served, including The Strip, Chinatown and other busy city neighborhoods. Vay‘s human remote drivers are located at its remote driving center.
Through the Vay app, customers can request an electric vehicle to be remotely delivered to their location within minutes. After the car has arrived, the customer takes control and drives it like a regular car. At the end of the trip, the customer exits the car and a remote driver takes over, eliminating the time-consuming search for parking. Agents work along with autonomous vehicle software to review and verify proposed routes when requested by the robotaxi when it encounters an environment it may not recognize.
Vay is also exploring use cases for the luxury OEM market. This will include tele-chauffeuring, where vehicle owners can drive themselves to a social event before being comfortably teledriven home. Luxury OEMs will also be able to offer pickup and drop off to charge points and maintenance for luxury vehicles. Vay’s offering will help these brands provide a bespoke experience to their clientele.
Nexar, a company delivering AI-powered mobility solutions, and Vay have formed a partnership aimed at integrating Nexar’s BADAS (Beyond ADAS) directly into Vay’s remotely driven fleet. The design partnership adds a proactive layer of safety as the service scales. Powered by Nexar’s Real-World Data Engine, BADAS learns from behavior captured across what is said to be one of the largest distributed vision networks on U.S. roads, which records more than 100 million miles of road data every month. For Vay, this means remote drivers gain augmented, AI-powered operational capabilities thanks to augmented safety features and vision-based data services – helping them anticipate risk earlier.
Low Latency
Soliton Systems specializes in low-latency encrypted video streaming technology, which is necessary for remote driving. The aim of its Zao products is to deliver ultra-low latency encrypted video streaming with full high-quality HD/4K video in a reliable manner, with a return path for remote control. Soliton claims its Zao technology can provide live video streams from the car back to the teledriver in less than 100msec. Other bonding systems typically have latencies of 1 to 2 seconds. Even over 4G, Soliton says its product can achieve latencies of 65ms and below, including both encoding and decoding. With 5G, this latency can be further reduced. It does this using bonded cellular connections and its own RASCOW2 (real-time auto speed control based on waterway-model) technology.
The Zao-X is a hardware transmitter for traditional broadcast and public safety events, and the software based Zao-SDK for Jetson is an OEM software library designed for the NVIDIA Jetson platforms. Soliton’s second generation codec for video optimization is a proprietary streaming protocol for unpredictable networks, such as 4G/5G cellular. It can optimize the video stream even if there is very low signal on all the network operators; ensuring video is delivered even in the most challenging network conditions. This technique, coupled with H.265 compression, provides a reliable stream even if some networks disappear or are in contention with many users, such as at a well-attended public event
As the Zao technology can provide two-way communication, it operates bi-directionally. Video is streaming from multiple cameras in the car In one direction from the car to the teledriver and in the second direction it is a secure IP network connection that sends the commands back from the teledriver to control the car. The Zao technology does not rely on lidar sensors. With their Zao products, Soliton is the first to market with a mobile H.265/HEVC encoder that could live stream full HD video over multiple 4G connections simultaneously. They also claim to be the first to offer a solution with a latency as low as 35msec from camera to the remote receiving platform.
Follow TTI, Inc. on LinkedIn for more news and market insights.
Statements of fact and opinions expressed in posts by contributors are the responsibility of the authors alone and do not imply an opinion of the officers or the representatives of TTI, Inc. or the TTI Family of Specialists.
Follow TTI, Inc. - Europe on LinkedIn for more news and market insights.
Statements of fact and opinions expressed in posts by contributors are the responsibility of the authors alone and do not imply an opinion of the officers or the representatives of TTI, Inc. or the TTI Family of Specialists.
