Deep R&D From University Lab On Route To Become The Technology Powerhouse In Japan

This is the second article in a series of four exclusive interviews with ACSL Ltd. from Japan. In this series, Mr. Chris Raabe, who is the Chief Technology Officer of ACSL, shared about their history of technology R&D and the latest cutting-edge technology to solve challenging issues in Japan.

The Need For Specialized Technology To Solve Specific Issues 

The earthquake and tsunami that hit Japan in 2011 left a devastating impact on the nuclear power plant in Fukushima. It required out-of-the-box and advanced technology to check the situation inside the nuclear power plant. There was one professor at the Electrical Engineering Department in Chiba University, Japan whose lab is the most advanced in autonomous control and autopilot technologies. The professor started his lab in 1998. After years of research and trailblazing the technology behind autonomous control systems, he successfully implemented his research in control algorithms for autonomous systems and applied it to single-rotor radio control helicopters. Additionally, he was also able to apply the control design to multicopters with single line LiDAR-SLAM (Light Detection and Ranging)- (Simultaneous Localization and Mapping) to do autonomous navigation. The achievements of his lab did not go unnoticed. The Japanese Government awarded his lab with a handsome grant to develop a specialized drone for the crucial mission of navigating inside the damaged Fukushima nuclear power plant. From there on, Professor Kenzo Nonami saw bigger potential of the technology in solving challenging and difficult issues. Therefore, he founded ACSL in 2013.

ACSL’s Drone Technology

Around the same time, Chris Raabe was at the University of Tokyo. Just like the research by Professor Nonami, his team was also developing autopilot. However, they were exploring autonomous navigation using Visual-SLAM instead of LiDAR-SLAM. Chris, who was from the Aerospace Department was collaborating with a researcher from the Information Science Department to combine their drone control and computer vision expertise to develop a drone that can navigate using computer vision instead of LiDAR. Although their vision-based implementation was not as mature as Professor Nonami’s LiDAR implementation, Chris believed that that was where the future was going with drone navigation. Eventually, Chris brought his expertise to ACSL when he joined them in 2017 and it turned out to be a major milestone for the company.

Visual-SLAM allows drones to navigate in a similar way that a human’s pair of eyes can. While Visual-SLAM provides robust, accurate results at a cheaper cost, LiDAR does the complete opposite. Nonetheless, LiDAR is useful in situations where visual information is difficult to get when drone is flying in a dark tunnel or places with very clean white walls. Instead of using machine learning or neural network for navigation, ACSL uses traditional image processing known as computational photography and mathematical geometry relationship to let drones construct, update, expand and reuse maps of unknown environments while keeping track of its goals, autonomously. The team is highly expert in programming a combination of micro controllers and GPU (Graphics Processing Unit) to ensure efficient image processing in real-time. Moreover, Visual-SLAM is also an alternative way to navigate in a non-GPS environment with an intrinsic kind of map that lives inside of the drone. However, ACSL does use Artificial Intelligence (AI) for safety measures especially when it involves drones flying nearby humans, Chris clarifies.

Continuous R&D To Give More Time For Humans 

In relations to safety measures, the R&D team at ACSL is focusing on two short term developments. First is related to providing anti-hacking, highly secured small drones for aerial photography that can be used with confidence by the local authority. Second is ensuring highly capable and safe drone delivery and logistics using simple multicopter drones with fixed wings. Chris adds, “We’re trying to maximize efficiency without sacrificing the simplicity of the multicopter mechanics. We think that is the best way to proceed with a safe and efficient logistics drone.”

For their medium term R&D, the focus will be to improve the usability of drones especially for the last mile delivery to home customers which is going to be a big focus in the near future. Generally, ACSL would like to provide a way for autonomous robotics to free humans from menial and dangerous tasks. By doing this, humans will be able to choose how to use their valuable time. This technology will also be introduced outside Japan in the near future.

Post-Pandemic Plan for ACSL

Before the COVID-19 pandemic, ACSL had big plans to expand to Southeast Asia with Singapore as its main base. However, that plan has been put on hold. This year, it is becoming more apparent that ACSL’s drone has dominated the Japanese market. Not resting on their laurels, during the period of the pandemic, ACSL prepared themselves and continued to expand the API (Application Programming Interface) for their drone so they can make custom applications that can work with third party systems and suppliers when the time comes. Towards the end, Chris gave his message to readers, “We think coming out of COVID-19, ACSL will have more momentum to expand outside Japan. I really look forward to work with companies and researchers in the ASEAN region.”

We hope you have gained greater insights after reading this article and looking forward to reading the third series in December 2021. Share your thoughts and feedback at info-asia@lne.st