In a recent online conference call, Herbert Diess, VW’s CEO, told his top managers that he believed the entire automotive industry was trailing far behind Tesla. In terms of software development, he believes that catching up won’t be impossible, albeit a particularly daunting task.
We’ve known about Tesla’s advantage in self-driving AI for ages now. In their early days, they competed with the likes of Waymo and Uber. They had been logging millions of miles on their training algorithms between their limited, technician-observed, fleets. However, with the massive adoption of Tesla’s lineup, they have blasted past the competition in terms of miles. Their unique neural network, which uploads error reports while running in the background of the consumer fleet, recently passed 3 BILLION miles.
What worries me the most is the capabilities in the assistance systems. 500,000 Teslas function as a neural network that continuously collects data and provides the customer a new driving experience every 14 days with improved properties. No other automobile manufacturer can do that today.
In terms of software, specialization is key, and having a decade of experience addressing the hardest problems in automation, can’t be quickly replicated by traditional automakers. If Volkswagen starts now, with the creation of a new division focussed on chasing Tesla’s coattails, then it’s not soon enough.
Their ID3 and other upcoming EV releases, need to aggressively overcome the problems which Tesla’s array of sensors are rapidly polishing up and packaging for release. For a company which is largely appraised as a software firm, perhaps Tesla could monetize their data through some sort of licensing in the future.