Eye tracker project.

I mentioned this project in a reply to a comment here a while back. This Honda ST1100 is the platform for a prototype project. I was teaching an Advanced Rider Training course. There was also a cornering clinic for instructors being conducted while the regular students were in the classroom. One of the instructors brought this bike up to give it a run and charge the battery so to speak! Since it was available, I took the chance to snap a couple of photos. Since this wasn't an artistic endeavour I didn't worry about moving the gal's gear. Besides, I was rushed for time. My blog pictures are art in a hurry!

Educating people in any area is a complicated process. So this isn't meant to make the process sound simple. However, there's a basic flow. There's a distant point the student needs to reach. In order to get the student there the educator needs to know where the student is right now. That will determine the pathway to the target.

The Eye Tracker project was started to establish a process for evaluating where rider skills are at a given time. Since "head and eyes" are such a critical element in riding, the idea was to find out how well riders were doing here. We'd like to know how effectively new riders use this skill compared to more experienced riders, for example. We'd also like to know if there's any sort of baseline that goes with different experience levels. For example, would there be a "typical" level we'd expect to see at five years, or ten years, or whatever?

The system itself is pretty cool, if somewhat primitive. Contained within the wooden box where the trunk would be is a laptop loaded with some neat software. This software has been developed by an outside source who agreed to let us use it. There is also some auxilliary equipment in the lower part of the box and in the saddle bags.

A rider dons a pair of goggles under their helmet. The goggles contain a video camera and some sensors that are aimed at the rider's pupils. Once everything is adjusted and calibrated, the rider goes off and enjoys some time on the bike. In the last outing where we tried to get data on a variety of riders each was asked to follow a prescribed route. It was also planned to be long enough that the rider would soon forget about the cameras and revert to their natural patterns.

Once the data is collected, everything is synched by the software. A video playback is available. It shows the road in front of the rider. Superimposed on the video is a small dot. It shows exactly where the rider's pupils were focused. The longer the rider focused on a spot, the larger the dot grows. It's pretty fascinating.

It's revealing for the rider to see the video of their ride. As expected, newer riders tended to have less eye movement than experienced riders. On the other hand, those who had been riding longer didn't always have better performances with head and eyes. There were several riders of ten or more years riding time that were looking three painted lines ahead of the front wheel in curves. Which is not unexpected considering that the leading cause of rider fatalities in our state is failing to negotiate corners. Investigations of where the crashes happened shows the riders simply weren't looking far enough ahead. The riders committed without having all the information.

Like I said, this was a prototype. More of a feasibility study. We'd like to get funding to do a much larger study. It would be a tremendous tool in helping find out what riders need.

Something else just occurred to me. The bike looks a lot like a Good Humour ice cream rig, doesn't it? I'm keeping my options open for a second career. Maybe it's time to go hunt up Conchscooter in the Florida Keys. I'm sure he could point me to a place I could ride around and sell ice cream to the masses!

Miles and smiles,

Dan