Sign up for daily news updates from CleanTechnica on email. Or follow us on Google News!
A recent LinkedIn post (also available on X/Twitter) by Alex Roy raises some interesting questions about the future of autonomous vehicles. While most people are testing Teslas with “Supervised FSD” on challenging local loops, the former coast-to-cost speed champion decided it was better to go all-in and test autonomous vehicles going from New York to Los Angeles. Before I can discuss whether this is a good test for developing AV technology, I’ll have to cover some background.
Testing Endurance For ICE, Then People, Then EVs, & Now AVs
If you’re unfamiliar with Cannonball runs, or have only seen the movies, the key thing to learn about Cannonballing (setting NY-LA records) is that it has always served a serious purpose. Early on, in the days of Erwin “Cannonball” Baker, an automaker used the record to prove that crossing the continent in 11 days, 7 hours, and 15 minutes was faster and safer than taking horses or trains. The U.S. Military, in a convoy partially led by a young Dwight D. Eisenhower, tried something similar just a few years later in 1919, pushing the boundaries of automotive technology.
Later racers improved automotive technology, created storylines for popular films, and protested illogically low speed limits. After automotive technology got really good and speed limits improved, nobody seriously Cannonballed for decades. The modern era of racing gas cars across the country with advanced counter-police technology (that costs a pretty penny to outfit a car with) began when Alex Roy took his BMW across the country in record time in October 2006.
Setting overall speed records still continues, and with the proliferation of modern EVs, records for travel under electric and even solar power are now being set, with the best EVs starting to approach gas car times. Now, semi-autonomous records are being set, with the goal of testing fully autonomous systems once they are capable of going from coast to coast.
A Large, Measurable Challenge
The problem with most semi-autonomous vehicle testing now is that it’s largely anecdotal. There are some controlled courses where AVs are tested consistently over and over, but when you consider that autonomous vehicles run on machine learning, it’s possible for software to fit itself to perform amazingly on frequently tested courses and miss things everywhere else in the real world. By finding the fastest route across the country (which can vary depending on a number of factors), a Cannonball run is almost guaranteed to give an AV system a unique challenge every time but also with clear goals.
By measuring overall time, disengagements of several types, and other errors, systems can be compared to others and improvements can be tracked.
The Downside: Avoiding Surface Streets
The only real downside to this rigorous testing method is that this kind of AV test avoids testing on surface streets. Surface streets aren’t completely avoided, as there’s a need to leave Manhattan, navigate to charging stops, and arrive in Los Angeles. But, most people do a mix of driving that includes more surface streets and less highways.
All in all, I think AV Cannonballing is important, but should be paired with some kind of consistent city street test that helps measure that side of things.
Featured image by CleanTechnica.
Chip in a few dollars a month to help support independent cleantech coverage that helps to accelerate the cleantech revolution!
Have a tip for CleanTechnica? Want to advertise? Want to suggest a guest for our CleanTech Talk podcast? Contact us here.
Sign up for our daily newsletter for 15 new cleantech stories a day. Or sign up for our weekly one if daily is too frequent.
CleanTechnica uses affiliate links. See our policy here.
CleanTechnica’s Comment Policy
Source: Clean Technica