Back in Taiwan, the Human-computer Interaction and Imaging System Laboratory at the Department of Electronic and Computer Engineering at National Taiwan University of Science and Technology (Taiwan Tech HIS-Lab) is also working towards realizing self-driving technology. In fact, the team has been so successful in this field that it had taken part in various projects from the Ministry of Science and Technology (MOST) thanks to matchmaking from the Global Research & Industry Alliance (GLORIA), and the team has also had its research results validated at the Hutoushan Innovation Hub in Taoyuan. Yung-Yao Chen, director of the HIS-Lab, says that the team has been researching smart sensor technology for self-driving vehicles over the past few years. This includes RGB camera, far infrared (FIR) thermal imaging, light detection and ranging (Lidar) information, mmWave radar, etc. The team's aim is to achieve sensor fusion by combining input from various sensor technologies then process them through AI to balance out the strengths and weaknesses of each sensor. Dr. Chen is the recipient of the 2020 Ministry of Science and Technology Ta-You Wu Memorial Award and is the Co-director of the Taiwan Tech Smart Electric Vehicle (SEV) Research Center.

Developing a Taiwan traffic condition database to localize self-driving technology

In response to the booming development of self-driving technology, SAE International proposed a "Levels of Driving Automation" standard that divides self-driving cars from Level 0 to Level 5 to serve as a reference for government legislation and consumer purchase. On this scale, Level 0 is complete manual driving while Level 5 is fully automated. Currently, almost all of the self-driving technologies available on the market are Level 2, with only a few vehicles reaching Level 3.

"The goal of the laboratory right now is to develop self-driving technology that is on par with major international manufacturers to make Taiwan's industry competitive with them, which would be between Level 2 and Level 3," Chen explains. "We are currently working with Taiwanese enterprises on a project to fuse RGB cameras, thermal cameras, and radar point cloud information into a model to improve the shortcomings of single-sensor image models. Another recent focus of the lab is to establish a database of Taiwan's traffic conditions to make the self-driving technology developed by the lab more in line with real-world traffic conditions. We hope to be able to enter partnerships with car manufacturers in the future so that the traffic condition database and related technologies can have more practical applications."

Integrating university-wide research capacity to accelerate smart EV industry development

Besides self-driving cars, electric vehicles have also been a topic of intense interest of late. As governments in various countries start to implement gasoline vehicle obsolesce for environmental reasons, major car manufacturers are now rushing to develop electric vehicles. Taiwan Tech also recently established a university-level Smart Electric Vehicle Research Center staffed by the research team led by Liang-Kuang Chen from the Department of Mechanical Engineering and Yung-Yao Chen from the Department of Electronic and Computer Engineering. The center represents the collated effort of Taiwan Tech, which draws from a combination of more than a dozen professors from the College of Electrical Engineering and Computer Science, the College of Engineering, and the College of Applied Science and Technology. The center seeks to contribute to the development of Taiwan's electric vehicle industry by conducting research on key technologies, undertaking industry-academia collaboration projects, and cultivating professional talents.

Liang-Kuang Chen, director at the Taiwan Tech SEV Research Center, says that the driver, vehicle, and environment are all equally important in ensuring road safety in the era of smart electric vehicles. The College of Engineering will primarily focus on researching control components responsible for steering, throttling, and braking but will also conduct failure analysis and prevention on parts like the electric motor vehicle and the brake system. The aim is to enhance driving safety by warning the driver when a failure occurs and remind them to take the appropriate countermeasures.

However, one of the main reasons behind consumer hesitancy to adopt smart EVs is concern regarding the battery. Issues such as short battery life, battery safety, and inaccurate battery capacity estimates are all driving forces causing range anxiety for the driver. Yi-Feng Luo, assistant professor at the Department of Electrical Engineering, points out that the solution for all the aforementioned battery-related problems lies with the battery management system (BMS), and the team hopes to solve all three problems by developing a reliable BMS system. Simply put, the BMS system developed at Taiwan Tech will be able to detect the health condition of each battery pack and notify the owner to get the vehicle examined when it detects a risk of failure. In addition, the system will automatically skip faulty battery packs during charging and discharging as well as adjust mileage estimations accordingly to facilitate real-world driving scenarios.

Chun-Yu Hsiao, assistant professor at the Department of Electrical Engineering, says that the mileage of EVs is largely dependent on the performance of the electric motor. Hsiao's area of research includes high performance motor design, optimization analysis for magnetic circuitry, and automotive power/steering motor design. For example, his team is currently developing a new type of permanent magnet direct-drive motor with high efficiency, small size, and long life, which is expected to greatly improve operational efficiency and driving stability for smart electric vehicles.

With an extensive portfolio in R&D for self-driving vehicles and smart EVs, Taiwan Tech will no doubt prove itself a fortuitous aid for the Taiwanese industry through more and more industry-academia collaboration projects.