Power Electronics Group finalist in Google's Little Box Challenge

A team from ECE ILLINOIS’ Power Electronics group is one of 18 finalists for Google’s and IEEE Power Electronics Society’s Little Box Challenge.

Assistant Professor Robert Pilawa-Podgurski, along with a team of eight students, has been working on this project for a year. In late October, they will travel to Golden, Colorado, to the National Renewable Energy Laboratory to drop off their final design.

Among more than 100 entries, the team’s design was chosen, based on achieved power density and efficiency, as well as novelty of the idea. The team’s inverter will be vigorously tested for up to 100 hours at the facility. The winner will be announced in January 2016 and the winner will receive $1 million.

“It has generated a lot of excitement because we are going head-to-head against the best research teams in the world,” Pilawa said. “It’s kind of like the Olympics for power electronics.”

The goal of the competition is to stimulate research in DC-AC power converters, inverters, which are essential components in everything from renewable energy integration, grid storage, and electric transportation.

To encourage innovative solutions in this important area, Google and IEEE Power Electronics Society partnered to establish a technical competition with clearly defined specifications, so that power electronics teams can implement and test research ideas, and perform apples-to-apples comparison against other designs.

The requirements are to create a power inverter that is smaller and more compact than existing solutions. Currently, an inverter is the size of a cooler. The goal presented by Google and IEEE: make the inverter at maximum the size of a tablet. The Illinois team’s current prototype is less than half that size.

“We’re definitely in the running,” Pilawa said. “Many excellent academic research groups and companies are competing, but I think we have a very competitive design.”

He said the competition is a chance to show off what the Power Electronics group can do, and is also an opportunity to apply many of the research breakthroughs his group has developed over the last few years in a well-defined product, and benchmark it against other solutions.

One goal of the challenge, as team member and graduate student Shibin Qin said, was to make inverters easier to use. By having a small, efficient design, the inverter can be used as part of the original power source, he said. This means the power can be accessed more quickly and with less hassle, making alternative sources of power like solar panels and electric cars easier to use.

Since the team had previously not performed research in inverter design, the first several months were spent researching better solutions than the state-of-the art.

“Our research group has worked primarily on DC-DC converters in the past,” Pilawa said. “While we have achieved outstanding results in that area of power electronics, the DC-AC conversion to be performed as part of this challenge presents unique challenges. Many other teams in this competition have spent many years working in this specific area, so we were certainly at a disadvantage from the beginning. At the same time, this also gave us the opportunity to start with a clean slate, with a very different design than conventional solutions. While the strategy was risky, in the end our new concept has shown remarkable performance, so it was a good decision in hindsight.”

Pilawa said it was clear from the get go that this was not going to be an easy goal. They faced circuit, thermal, and packing challenges, along with learning to build a 2-kilowatt inverter that was smaller than an iPad.

One approach to this challenge is to carefully tune existing designs for this specific competition. However, this type of optimization is best done by industry, and does not play well to the strength of academic research, which should look for more radical approaches, Qin said. 

Therefore the team decided to go a different route, to really innovate on circuit and control design. In this way, the end-goal was not just to make a competitive entry to the challenge, but also to have a long-term research impact on the field of power electronics.

“We have learned things that were outside of our direct area of expertise,” Pilawa said. “And it made the students better engineers and they’ll be highly sought after by industry and academia.” 

It involves a higher level of perseverance, otherwise it really is not going to work, Qin said. There is always something you can do to make it better, it’s a never-ending process until the deadline, he said.

Yutian Lei, Christopher Barth, Wen-Chuen Liu, Andrew Stillwell, and Qin, who are all graduate students and Intae Moon, an undergraduate, all worked on the project for portions of the past year. Since September of this year, graduate students Thomas Foulkes, Derek Chou, Zitao Liao, and Zichao Ye have also helped with improvement to the final prototype. They’ve all put in a tremendous amount of work, Pilawa said, even while managing their coursework and other responsibilities.

Many competing teams have put tremendous financial resources behind their design, with large numbers of full-time engineers dedicated to this competition, Pilawa said.

“This was a nights and weekends project, as the students were also performing research on their primary research projects, which were generally not related to this competition. We’ve done it with graduate and undergraduate students, and we have done it on a shoestring budget,” he said. “No matter what happens in the end, I am very proud that under those conditions, we are in the top 18.”