One of the biggest challenges with data analysis - and I knew this from previous experience - is simply the amount of time required to sift through the mountain of data compiled during each session. There is only so much time available between on-track sessions, and it's practically impossible to go over every lap and every channel trying to find useful information that can help the team move forward. This is always an issue, and I certainly felt the added pressure at Mosport (CTMP) with John and Jodi looking over my shoulder!
Part of the solution is to make additional math channels or use certain graphical methods that make it easier to quickly and easily see exactly what you want to see. For example, one way to evaluate how the traction control system is working is to look at rear wheel speed compared to GPS speed, which will indicate how much wheelspin is occurring. The team's Bazzaz system utilizes a programmable throttle-position/rpm array to change its operation, and it also outputs a signal when it is active (which we feed into the AiM system). Put all those channels - GPS speed, wheel speed, throttle position, rpm and TC active - on a standard graph vs. time or distance, and it is practically impossible to make any sense of it - as you can see from the attached image.
A math channel showing actual slip (by mathematically comparing wheel speed to GPS speed) eliminates one data trace, and some creative charts can be used to eliminate others. If we are interested in how slip varies with throttle position, for example, it makes sense to show it just like that on an X-Y chart rather than showing both channels against speed or time. The AiM software includes many options for graphing data, including the X-Y plot; in this manner, a quick overview of the traction control system is available, and areas requiring further investigation can be easily seen.
The other part of the time-crunch solution is having a structured plan after each on-track session. In one of the books I have, <i>A Practical Guide to Race Car Data Analysis</i>, author Bob Knox outlines a procedure to that effect, and even includes a checklist that can be used; I realized at CTMP just how important this is. In a nutshell, certain data channels must be at least checked after every session - for instance, ensuring that the engine is not over-revving anywhere on the track by a quick look at the rpm channel, or checking that the fork or shock are not bottoming.
Once the basic checklist is complete, attention can be focused on what the rider and crew chief are interested in - at Mosport, as I mentioned, it was usually the suspension and its related channels. Once a decision has been made for setup changes and a direction for the next session, focus can then shift to the remaining data channels and finding any further information that can help the team.
The races at CTMP did not go as well for us as the final at Atlantic Motorsport Park. While Jodi was strong in practice in qualifying, he crashed out of the lead group just a couple of laps from the finish of Saturday's race. On Sunday, a fairing mount broke but we also went astray on the setup from Saturday, holding Jodi back to an eighth-place finish.
As with anything, putting theory into practice is not easy; the data is a big help to setup and provides many answers, but compromises must still be made. And, of course, we are still learning how to best use the data.
Since the races at Canadian Tire Motorsport Park, I have been working on combining the suspension channels with other data in order to better see certain parameters that we have discovered to be important. The next event for the team - and another chance to put theory into practice - is the final round of this year's Canadian Superbike Championship on August 25 at Le Circuit Mont-Tremblant.