The news is full of self-driving cars and while there is some bad news, most of it is pretty positive. It seems a foregone conclusion that it is just a matter of time before calling for an Uber doesn’t involve another person. But according to a recent article, [Ernst Dickmanns] — a German aerospace engineer — built three autonomous vehicles starting in 1986 and culminating with on-the-road demonstrations in 1994 for Daimler.
It is hard to imagine what had to take place to get a self-driving car in 1986. The article asserts that you need computer analysis of video at 10 frames a second minimum. In the 1980s doing a single frame in 10 minutes was considered an accomplishment. [Dickmanns’] vehicles borrowed tricks from how humans drive. They focused on a small area at any one moment and tried to ignore things that were not relevant.
Not that there were not problems. The software had some sort of learning algorithm, but it had trouble dealing with things like washed out road markings. However, he would go on to work with the United States’ Army Research Lab and that work would go on to inspire DARPA autonomous vehicle challenges.
One of the quotes in the article about current self-driving cars really struck home with us:
Virginia Dignum, a professor at the University of Delft, agreed that if AI researchers keep focusing primarily on deep learning, “at some point, people will be disappointed.” The field, she said, has to look beyond it and invest in other approaches…
We’ve often thought that the more software looks like a human’s brain, the more it will suffer from all the things we don’t like about human brains. The article isn’t terribly technical, but it does have an interesting bit about how current approaches mostly focus on vision that confirms what it already knows to be in a certain area. This is good for computing power, but not great for unknown roads and changing conditions.
[Dickmanns] wanted dynamic vision that can adapt to driving in areas that are not already carefully mapped or have changed significantly. This would be especially important in military applications where you might have to go somewhere unusual, an adversary might change things to confuse you, or roads could be damaged. [Dickmanns] has his own website discussing the technology and you can see a video of his work on that site. There’s also a wealth of more technical data on his three generations of computing platforms, starting with a description of the first system that processed 32×32 pixels of data at a time with an 8085 CPU, to the shift to 80×86, Transputers (45 of them), and — eventually — to dual-processor PCs. If you want to know that technical details, that’s the site you want. You might also enjoy the short video clip describing the milestones that [Dickmanns] accomplished.
Image credit: [Ernst Dickmanns] CC By-SA 3.0