Tokyo, Japan – Replacing human drivers with AI technology in autonomous racing vehicles may seem like an easy fix to alleviate the challenges faced by teams and owners. However, experts are now weighing in on the limitations of this approach.
Giovanni Pau, Team Principal of TII Racing, notes that human intuition plays a significant role in driving, particularly when it comes to estimating tire grip. “It’s impossible today to do a correct grip estimation,” he says. “A thing my friend Daniil (Kvyat) can do in a nanosecond.”
While the Technology Innovation Institute (TII) develops the hardware and software stack for all vehicles, teams are responsible for building their own system on TII’s platform. This approach raises questions about whether AI is truly capable of learning and adapting to new situations.
In reality, most so-called “AI” technology in racing is still programmed responses to an environment, rather than true learning. The recent demonstration race against former F1 driver Daniil Kvyat at the Suzuka Circuit in Japan highlighted this point. Despite the use of advanced sensors and computers, the vehicles were unable to swerve back and forth to warm up their tires.
As teams navigate the complexities of autonomous racing, it’s clear that replacing human drivers with AI technology is not a simple solution. The challenges of the sport remain unchanged, and experts are now questioning whether this approach truly addresses them.
Source: https://arstechnica.com/cars/2024/12/man-vs-ai-race-scrapped-after-ai-car-crashes-into-wall-on-warm-up-lap