Using Data Acquisition

Like most electronics-based technology, data acquisition systems have become increasingly sophisticated yet less expensive over the past couple of decades. It wasn't too long ago that data setups were only found on top factory racebikes, but even elaborate systems are now within reach of club racers and trackday riders. Data acquisition has even made it to production bikes, with a four-channel system as standard equipment on Ducati's 1098S.

A more traditional system like this Aimsports setup uses a wheel sensor to record speed and measure distance from a beacon. On-board accelerometers and a gyroscope can be used to create a track map, with position calculated by distance traveled from the beacon.

The basic premise of any motorcycle data acquisition system is that important information-such as speed, throttle position, rpm and so on-is recorded while the bike is in action, and data can be used to better evaluate the bike's (or rider's!) performance. Just as there is practically an infinite amount of data that can be recorded, there are numerous ways to interpret that data. In this installment of Art & Science, we'll discuss the important aspects of a basic system and how to begin collecting and analyzing data. In upcoming installments, we'll explore more sensors and other ways in which the data can be used.

The root of all setups is a single channel that records the bike's speed a set number of times every second. Speed is generally considered the base channel and one to which many other channels are referenced. The data, saved in memory, is downloaded to a computer that displays the information in graphical format as speed versus time or distance. Even this single channel can tell a lot about what's happening on the track, but collecting data from more channels and sensors obviously allows a more in-depth look at overall performance.

Speed can be recorded in one of two ways. The first is via a magnet or sensor on the front or rear wheel of the bike. The advantage of this is simplicity: The sensor is inexpensive, and often the bike's speedometer sensor can be tapped into with a single wire. From this information, distance can also be calculated; with a beacon at a set point on the track establishing a reference point, the bike's location on the track can then be found. Because motorcycles lean and the edge of a tire has a smaller circumference than the center, however, speed and distance measured in this manner change with lean angle and wheelspin, adding a margin of error.

A conventional system requires (at a minimum) a wheel-speed channel, which can be hooked to a sensor or tapped into a bike's speedometer harness. Tire circumference (and potentially gearing) must be noted and taken into account.

The second way in which speed (and position) can be recorded is with a GPS (Global Positioning System) receiver. With the electronics required for GPS becoming increasingly compact and cost-effective, more data acquisition systems are utilizing this method to measure speed and distance as opposed to using a wheel sensor. GPS, in combination with on-board accelerometers, can be quite accurate with none of the problems associated with changing tire circumferences. Additionally, there is no need for a beacon to establish a reference point.

Even by recording only speed, we can tell a lot about what's happening on the track. Most software packages allow data from two laps or sessions to be overlaid in graphical format, and this is the easiest way to highlight differences between any two data streams. For example, speed over the course of a lap for two riders can be used to quickly see where on the track each rider can improve. Or speed for the same rider over two different laps can be compared to observe changes over time.

Here is a wheel speed vs. distance plot for two laps with the same rider from California Speedway, using data from an Aimsports system. There are small differences between the two laps, pointing to areas that may need attention, either in bike setup or riding. With only speed traces and no data from a second rider to use for comparison purposes, it's a matter of working with the rider to find what changed from lap to lap and why. Note the small fluctuations in chicanes where the bike is turned and the tire's circumference changes, influencing the measurement.

This graph of wheel speed for two riders at a braking marker shows some interesting detail where the riders go from accelerating to braking. Rider A's trace (red) shows a crisp, sharply marked switch from acceleration to deceleration, whereas rider B's line (blue) is slightly curved, indicating a more leisurely transition and lost time. Additional data from throttle position or brake pressure would provide more information from which to draw a conclusion and suggest a change, but even the speed traces alone show rider B could benefit from some attention in this area.

This plot shows GPS speed for two different riders through a switchback series at Autobahn Country Club. The speed traces cross each other, indicating each rider's different approach to the section; how do you tell which way is quicker? Most systems' software packages allow the track map to be broken down into sections, with split times displayed for each. Here, rider A (red) is slightly quicker with an 18.76-second time vs. 18.88 seconds for rider B (blue). GPS setups use imaginary lines drawn perpendicular to the direction of travel through an exact point on the ground to mark the beginning and end of a section, an accurate method. Use discretion when looking at split times with non-GPS-based systems, however; because the section start and end points are calculated using distance from a beacon, they may be quite different from one lap to the next. In either case, avoid putting much stock in the "theoretical best lap time" that most software packages calculate by summing the quickest section times from a series of laps.

The trouble is, of course, that many people will not willingly share their data. Or you may go to a new track that you don't have previous data for, leaving you with none for comparison purposes. In this case, more channels are definitely beneficial, as they will allow you to draw conclusions and make changes based on characteristics that are similar from track to track or rider to rider. In upcoming issues, we'll look into some of these other channels in more detail.