TuSimple’s Autonomous Freight Network (AFN) is poised to become a new freight transportation backbone for the supply chain, enabling goods to be transported more safely, reliably, and efficiently than ever before. A key part of the AFN is a network of high-definition digital maps that TuSimple’s autonomous trucks use in combination with onboard sensors to perceive the road and safely navigate. TuSimple maps are highly sophisticated and include layers of rich data that enhance safety for everyone on the road. Now, TuSimple is unveiling a new dynamic map that creates a network effect for safety, called TuSimple CyberMap.
Each autonomous truck operating on TuSimple’s AFN perceives its surrounding environment with HD Cameras, LiDAR and Radar; however, the perception system alone can’t see several minutes into the future, around a long corner or through dense traffic that is obstructing its view. TuSimple CyberMap can help increase safety in these and other similar scenarios. TuSimple CyberMap brings a paradigm shift in safety capability by continuously mapping the TuSimple AFN, sharing information between trucks in near real-time, and adding a new map layer that captures information that is temporarily part of the road environment but might change in the future. For example, during a freight delivery mission, one truck might detect a car in the emergency lane, a pedestrian, or debris on the road. With this new knowledge, other trucks in the area can then alter their paths to the safest possible one.
One of TuSimple’s trucks perceives an accident on the side of the road with emergency-lane vehicles and road debris. That information is sent back to TuSimple’s data center. Once the data is processed, all other trucks operating on the AFN in the same region, pull that data and fuse it with their existing knowledge of the road and surrounding environment.
Trucks that have not yet passed the accident can alter their motion planning based on this new knowledge to maximize safety, for example changing lanes to avoid debris and emergency lane vehicles. This entire process can happen in as little as a few seconds.
Once one truck perceives that the accident and emergency-lane vehicles are no longer present, the same process occurs to notify the data center that the accident is gone, which is then pulled in by all trucks operating in the same region, and they no longer plan in response to that accident.
TuSimple CyberMap increases the safety of each truck operating on the AFN by leveraging information dynamically acquired and shared by all trucks operating in the same region. If one truck sees a hazard, in near real-time, each other truck will know how to best navigate that potential hazard, even if it is miles away.
Each additional truck operating on the AFN, in the same time window, increases the frequency of detection of potential hazards. The more trucks that are operating on the AFN, the more up-to-date the map is as it relates to potential hazards. This network effect increases safety with each autonomous truck added to the network.
Our goal is to improve transportation safety for all road users by delivering the safest driver on the road. TuSimple CyberMap’s capability to dynamically alert other trucks about upcoming potential hazards is one of the ways we are doing so. Ultimately, through safe engineering, testing, and deployment, we have the opportunity to save thousands of lives every year.