KUT team places fourth in international FPGA design contest aiming at fully autonomous driving
The 2019 International Conference on Field-Programmable Technology was held in Tianjin, China from December 9 to 13, and the team from KUT's Integrated Systems Laboratory (led by Associate Professor Yukio Mitsuyama) won fourth place in the Field Programmable Gate Array (FPGA) design contest in which the contestants designed the structure of integrated circuits.
Team members: front row, from left: Itsuki Ikeno, Tomonari Tanaka, Riku Tsuruoka and Takumi Kuchiba (all fourth-year undergraduate students); back row, from left: Associate Professor Miyama and Research Associate Dr. Wang Liao.
Field programmable technology, in which designers can freely write code to design logic circuits, has applications in a wide range of fields, including high-performance computing systems, low-powered control systems and mobile communication devices. Recently, it has also been employed for implementations of embedded machine learning applications--so it is attracting attention as a key technology for realizing the Internet of Things (IoT) society of the future.
The design contest theme was "FPGA technology required for future level 5 autonomous driving (autonomous fully automatic driving)." The contest task was to develop an autonomous vehicle controlled by an FPGA with the closest to perfect autonomous driving. On miniature roads set up at the venue, the participants competed to create the most accurate travel on a given route, with recognition of features such as pedestrian crossings, intersections, traffic lights, obstacles and pedestrians.
Fourteen teams from three countries participated. The KUT team placed third in the preliminary round with its stable driving, achieved by means of multiple cameras and FPGA boards, and the cooperative operation of hardware processing and software processing. The team performed very well in the final round too, placing fourth even though it was their debut in the contest.
The development of an autonomous vehicle using FPGA requires not only FPGA design skill for implementation of image recognition processing, but also knowledge and technical skill related to information, machinery and electronics, ranging from vehicle body design to electronic work, so the degree of difficulty of the contest is extremely high. Most of the participating team members were graduate students, but the KUT team members were all undergraduate students; they were highly evaluated by the other participating teams.
Mr. Tanaka, the team representative, said, "The FPGA design contest was a valuable opportunity for us to interact with people from many universities, exchange technical opinions, and get inspired by the perfection and technical skills of the higher-level universities. Our result was a bit disappointing, since we couldn't finish the design and implementation of all the functions that we had planned in time for the contest. We will do our best to apply that experience in our future work."
- Professor Fujita's research team opens the way to high efficiency thermoelectric conversion devices
- 2020 Feldman Prize goes to Associate Professor Yutaka Kobayashi for his mathematical analysis of a cultural evolution model
- Research group led by Prof. Fujita develops new catalyst for converting greenhouse gases to hydrogen and chemical raw materials by light irradiation