Chen places second at Boeing IT Case Competition

4/10/2015 Katie Carr, Communications Coordinator, Coordinated Science Laboratory

AE grad student Derek Chen takes second in Boeing competition.

Written by Katie Carr, Communications Coordinator, Coordinated Science Laboratory

Aerospace Engineering graduate student Derek Chen won second place at Boeing's IT Case Competition earlier this month. Left to right: Illinois undergraduate student Anselmo Shim, CSL professor Grace Gao and Chen.
Aerospace Engineering graduate student Derek Chen won second place at Boeing's IT Case Competition earlier this month. Left to right: Illinois undergraduate student Anselmo Shim, CSL professor Grace Gao and Chen.
Aerospace Engineering graduate student Derek Chen won second place at Boeing's IT Case Competition earlier this month. Left to right: Illinois undergraduate student Anselmo Shim, CSL professor Grace Gao and Chen.
Derek Chen, a first year masters student in Aerospace Engineering at Illinois, was recently awarded second place in Boeing’s IT Case Competition, a yearly intercollegiate competition that showcases student IT talent organized by Boeing.

Students were asked to recognize and identify 3D objects within a 3D Light Detection and Ranging (LIDAR) scan of Mt. Rainier and were judged on accuracy, efficiency and clarity of the program they wrote. LIDAR data is a laser scan of a particular landscape represented as point cloud data in 3D space. Participants were given five 3D models of various objects scattered randomly across the landscape and required to develop an application that could decipher between the 3D models and the landscape point cloud and identify the location of the objects.

“Nowadays, there’s so much data because it’s so convenient to just take a picture or video of an area,” Chen’s advisor and Aerospace Assistant Professor Grace Xingxin Gao said. “But what people really care about though is not just the image, but being able to automatically identify areas of interest and identify shapes.”

Chen added that this type of competition is important to Boeing because they often collect data like this, but need the ability to quickly identify common shapes, or shapes obstructed by the terrain, such as silos, skyscrapers or military tanks.

As second-place winners of the competition, Chen and Gao, along with undergraduate student Anselmo Shim, were invited out to Boeing’s headquarters for the award’s ceremony earlier this month where they received a tour of the Boeing Everett Factory outside of Seattle and a cash prize of $1,000.

Chen took a unique approach to solving the problem by matching 3D points into an image grid that allowed them to process and manage the 3D points, enabling them to search for 3D points in a nearby space in a quick manner.

“The 3D data they provided us was very unordered,” Chen said. “The most common method to solve a problem like this is to do template matching, where you try to fit data points to a 3D model to see if there’s a 3D object there. However, given the large dataset provided, doing something like that would have taken way too long.”

Chen’s approach allowed them to search for nearby points and identify the 3D objects in a much quicker, less computationally expensive and more effective manner.

“We were able to process the data points much faster and obtain a filtering algorithm to find most likely points within the millions of points they provided us,” he said. “Instead of matching for an entire 3D model, we took specific descriptors of the shapes they wanted us to find to see if we could find a match.”

Gao added that this problem was very difficult and gave Chen training in how to tackle an open-ended, real-world problem.

“Unlike homework assignments or exam questions, most real-world problems aren’t well-defined, so you have to take initiative and try to explore and try to find a solution and if there even is a solution,” Gao said. “This award should be a big encouragement for him to continue innovating and progressing in research.”

Chen works with Gao on research related to GPS/GNSS-based positioning and navigating and timing for applications such as UAVs, power systems and robotics.

“Our research is on GPS with sensor fusion for localization,” Chen said. “With the experience of processing LIDAR data through this Boeing IT competition, I could possibly use it for navigational purposes and to improve positioning in my current research.”

 


Share this story

This story was published April 10, 2015.