Artificial intelligence might be smart enough to beat champions at chess, and outclass many world class players in their own game. One thing they can't do: Defeat the world's fastest cyclist for the speed record of the fastest human-powered vehicle at 83.13mph. Well, that would require a human.
That being said, it doesn't necessarily mean that AI can be of no use to design the world's most streamlined bike by providing the necessary measurements that would allow the bike to cut air resistance to an absolute minimum. Researchers from the Computer Vision Laboratory of Switzerland's École Polytechnique Fédérale de Lausanne (EPFL) have developed a new machine learning algorithm for the same application.
According to the creators, the algorithm was trained on the aerodynamic qualities of different 3D shapes, which makes it capable to understand the laws of physics.
"Instead of computing solutions of equations or simulating moving particles, our algorithms predict the aerodynamic performances from previous experience, the same way a human engineer would do," researcher Pierre Baqué told Digital Trends. "By doing so, we reduce the time to estimate the performance of a new design from several hours to a few milliseconds, which lets us implement computer-based automatic shape optimization."
The new technology which is called Neural Concept is used as the starting point for a newly designed "aero speed bicycle," built at France's Annecy University Institute of Technology. The bike will be used to attempt the World Human-Powered Speed Challenge in Nevada this September after it has its first trials in the coming days.
The current record holder is Sebastiaan Bowier's rocket-like bike. It will be interesting to see if the Neural Concept's algorithm comes in handy to break the record. Baque has noted that sometimes the measurements are 5 to 20 percent more aerodynamic than conventional methods.
As for the future of Neural Concept? "Our startup is developing commercial applications of the technologies for generative design, based on deep learning," Baqué said. "We are starting collaborations with industrials, while continuously developing our software."
If Neural Concept does break the record, it will be a new achievement for developers and will also open new doors for the use of artificial intelligence.