Using the Microsoft Xbox Kinect Sensor™ for Child Safety Research
Aditya Belwadi, PhD
March 5, 2015
Installation of "virtual rear-facing child seat" in a sub-compact sedan.
It is no surprise that researchers often seek out cost-effective ways to conduct rigorous research. What may be surprising, however, is that a new data collection tool being utilized by CIRP@CHOP investigators is part of a gaming system that many families have in their homes.
Several lines of research supported by the Center for Child Injury Prevention Studies (CChIPS) are currently leveraging the capabilities of the Microsoft Xbox Kinect Sensor™.
A line of research that I am leading uses the Kinect for Windows sensor to create a “virtual child seat” of rear-facing child restraint systems (CRS). The digital model is a representative volume of all rear-facing infant seats in the US market (as of April 2014) to aid vehicle and child seat manufacturers in assessing space and fitment during the design phase of the vehicle rather than assessing fit after both the vehicle and child seat are on the market, which has typically been the process.
The current industry standard for this type of scanning utilizes a laser scanner and/or digitizers. Although these scanners provide high accuracy and resolution, the associated time, cost and software are substantial. Alternatively, we used the Kinect for Windows sensor to scan 72 rear-facing CRS, forward-facing CRS, high-back boosters, and low-back boosters that represented 252 commercially available CRS.
Each seat took approximately 2 hours to scan using the Kinect sensor at a cost roughly 100 times less than traditional approaches without sacrificing the level of detail and accuracy necessary for this project. This line of research will be presented at the SAE World Congress and Exhibition next month in Detroit, Michigan.
The capabilities of this technology go beyond scanning inanimate objects like CRS. It is well known that the Kinect sensor tracks participant movement during video game play. In another line of CChIPS research led by the Center's Co-director Kristy Arbogast, PhD, the Kinect sensor is being used to determine the positions child occupants assume while riding restrained in cars. The results from this research can ultimately help to improve testing conditions for vehicle and restraint safety devices.
As we begin to utilize “unconventional” data collection tools such as the Kinect sensor for scanning, tracking, and quantifying shapes of safety devices and actual people, we are seeing that the technology is very promising for several applications beyond the gaming industry.
Aditya Belwadi, PhD is a Research Scientist I for biomechanics research at the Center for Injury Research and Prevention at The Children’s Hospital of Philadelphia (CHOP).