Comparing headlight configurations
for visibility
How would you address the demand for motorcycle safety as an OEM (Original Equipment Manufacturer)? How would you evaluate motorcycle visibility for future production models?
First steps:
Tasked with evaluating two experimental motorcycle headlight configurations, the focus for this study was improving visibility and safety of motorcyclists on the road. Two months were dedicated to creating the test plan, and identify how best to conduct the evaluation. A lot of question needed to be answer before this study could start.
Scenario
The first question was how the headlights should be compared. After looking into motorcycle accident research, we determined a consistently dangerous scenario was when a car made a left turn in front of a motorcycle.
Location
The next question to answer was where. While most of our studies took place in a high-fidelity driving simulator, to be able to properly evaluate the headlights, we needed real world lighting. A private test track on a closed course was chosen so that we would have full control of the situation. This also allowed us to have an elaborate setup.
Performance measurement
After the determination of the scenario and the location, we needed to figure out how to evaluate how visible the different headlights were to drivers. Since the drivers were making a left had turn, we needed a way to measure whether it was a successful left turn and how safe that turn would be. After a lot of discussion and trial and error, we worked out a method using GPS to synchronize tracking of vehicles with the drivers behavior.
Team
The final step was bringing on the necessary personnel. I oversaw a team of:
-
2 motorcycle riders,
-
3 car drivers, and
-
2 junior research assistants.
Full-scale testing:
Participants
Typical drivers were identified as the target demographic.
Recruitment was done in-house with word of mouth responsible for finding the 24 participants. I worked closely with a colleague to coordinate the participants.
Sessions
Each session included an introduction, calibration, driving tasks, and a dismissal phase. Sessions lasted up to three hours each night with no more than two participants per night.
Driving task
Drivers sat in a car positioned at a makeshift intersection, with the car immobilized for safety. A platoon of three vehicles would approach from the opposing lane. As the platoon passed, the driver made a decision about whether or not they could safely make a left turn. When the judged it was safe, they pressed the gas, and that information was recorded.
This task was repeated so participants experienced variations of:
-
gaps between the vehicles, or
-
whether the middle vehicle was either one of the motorcycle headlight configurations or a car.
Bringing the data together:
Due to the sheer volume of data and the importance of the synchronization, I spent a month overseeing the validation of the data. A series of ANOVAs was used to determine which configuration resulted in the largest acceptable gap.
A comprehensive report detailing the study, results and recommendations, was delivered to the client.
