Invited TALK

AI Robot Competitions vs Motorsport :　Parallels in Driving Technological Innovation

Abstract:

The history of the automotive industry provides a powerful perspective for understanding current trends in AI-driven robotics. From the mass production of the Ford Model T to high-performance innovations driven by motorsports, automotive development has followed a clear pattern: rapid prototyping under harsh conditions through racing, standardization of parts and establishment of safety regulations, and large-scale industrialization.This keynote speech will highlight the similarities between this industrialization process and the emerging field of AI robotics, with a particular focus on the role of robot competitions as catalysts for technological innovation.

　Competitions such as RoboCup@Home, the DARPA Robotics Challenge, and the Food Topping Challenge will be analyzed for their similarities to automobile racing in terms of their function as test beds for verifying reliability, performance, and safety under realistic or extreme conditions. By comparing the timelines, technological breakthroughs, and evolution of academic curricula in both fields, we will identify common mechanisms by which competitive environments accelerate innovation.

The presentation will highlight key similarities and differences between the two industries as they transition from basic research to sustainable, standardized deployment. It will also clarify the importance of academia's role in promoting high-risk, high-return research, fostering industry collaboration, and cultivating the next generation of engineers.

Field Robotics – Agriculture and Construction Sites

Abstract:

  The social demand for field robots that operate outdoors continues to increase. While many robots are already functioning in the field, there remain numerous tasks that still rely heavily on human labor. In this presentation, two examples of field robots will be introduced: crop-harvesting robots designed to support human manual work, and construction robots equipped with force control capabilities.

For crop-harvesting robots, we first discuss the working environment and the technologies required to operate effectively. For construction robots, we analyze the necessary forces and present an excavator equipped with a Series Elastic Actuator (SEA) to achieve these requirements. Finally, we will discuss the essential requirements for robots to be successfully deployed in field environments.

Artificial Vision: How Can Computer Provide Useful Vision with the Brain? 

Abstract:

Human perceives meaningful event, or affordance, in real-world to take action　with unique and characteristic time course, and therefore robot also should do for natural and seamless interaction with human. In human visual system, the retina plays important role in dynamical computation and encoding of the meaningful event to send it to the brain. In the past, the function of the retina had been greatly underestimated. Recent physiological studies, however, have unveiled several lines of new aspects concerning more sophisticated function of the retina in visual perception. In my talk, I will first review this stream of the past and the recent retinal studies and explain how the retinal circuit effectively and efficiently computes and encodes image in real-time to communicate with the brain, referring to relevant parallel neuronal circuit architecture that is thought to be critical to dynamically evoke perception. Next, I will demonstrate a real time event-driven image computation/encoding technique using the integrated circuit vision sensor with analog/digital mixed architecture, namely neuromorphic retina, devised in our lab, hoping to foster further development of neuromorphic sensory systems to provide the embodiment with human-friendly robot. Finally, I will envision possible future application of the neuromorphic retina to brain computer interface.