A recently published Tesla patent application titled “Autonomous Driving System Emergency Signaling” describes a method of quickly communicating emergency information from vehicle sensors feeding into autonomous driving software. The new communication method will improve Autopilot’s response in emergency situations, thereby reducing the probability of accidents.
Tesla’s invention takes latency in data transmission into account as an area of improvement. In general, critical information can get stuck waiting to be processed by a computer after non-critical information that’s ahead of it. Under Tesla’s US Patent Application No. 2019/0138018, critical emergency situations detected by sensors are moved to the front of the line for priority processing and response. Tesla’s invention achieves this using two main approaches.
First, the transmission from sensors that detect an emergency sends their findings to the main computer at a higher transmit power than other messages. Other signals at lower power transmissions are then interpreted as ‘background noise’ compared to the emergency signal. This process is described in the patent application as follows:
“When an autonomous driving emergency event is detected by an autonomous driving sensor…the [sensor] transmits the autonomous driving emergency message in a non-assigned time slot at a higher transmit power level than a transmit power level of an autonomous driving sensor…Because the autonomous driving emergency message is transmitted at a higher power level than the transmission from the autonomous driving sensor, the transmission from the autonomous driving sensor may be treated as background noise by the autonomous driving controller to thereby receive and decode of the autonomous driving emergency message.”
In a second approach, the autonomous driving sensors that encounter an emergency message are programmed to stop sending signals, and the vehicle’s main computer will direct them to resume communications after receiving the emergency message. This process is described in the patent as follows:
“…if an emergency transmission is detected…the autonomous driving sensor ceases transmitting autonomous driving data. Such cessation may continue for one assigned time slot, for more than one assigned slots, and/or until the autonomous driving sensor receives direction from the autonomous driving controller to continue transmitting autonomous driving data or receives a new…bus time slot assignment from the autonomous driving controller. During this time period…the autonomous driving sensor continues to collect and buffer autonomous driving data.”
Several variations of achieving these two main concepts are also described in the application and invention claims, including managing the specifics of the transmit power level differences and reassigning time slots for sensors to communicate on the data bus. Overall, this recent patent application is yet another indicator of Tesla’s continued improvement of its autonomous driving capabilities.
Tesla’s advances in the autonomous driving arena have been touted by CEO Elon Musk and industry experts alike. ARK Invest analyst James Wang recently estimated that the all-electric car maker’s decision to develop its Full Self-Driving computer chip in-house put the company four years ahead of the competition. Musk, for his part, declared the chip the best in the world at Tesla’s Investor Autonomy Day. “It seems improbable. How could it be that Tesla, who has never designed a chip before, would design the best chip in the world? But that is objectively what has occurred,” Elon touted.
While Tesla has yet to roll out the total capabilities of its Full Self-Driving suite, Musk has said on several occasions that the software will be “feature complete” by the end of 2019 with only regulatory hurdles left for full release.