We have been working on the large project on AI and Robot in the Moonshot program since 2020. Focusing on the coevolution and self organization capabilities, based on the Society 5.0 concept, it is a new and challenging program aiming at the AI robotic system in 2050, which will contribute on solving some of Mega-trend problem in our society. I will introduce some of the projects in this program for realization of the Society 5.0 by back-casting technologies from the 2050 to the current ones. Then I will show the progress and current status of several projects of the Program since there are new projects on Science of Awareness and infrastructure construction on the moon by AI and robotic technology.

Toshio Fukuda is Professor Emeritus of Nagoya University and University Professor Waseda University. He is mainly engaging in the research fields of intelligent robotic system, micro and nano robotics, bio-robotic system and industry applications in robotics and automation. He was the President of IEEE Robotics and Automation Society (1998-1999), and IEEE President (2020). He was Editor-in-Chief of IEEE/ASME Trans. Mechatronics (2000-2002). He was chairs of many conferences, such as the Founding General Chair of IEEE International Conference on Intelligent Robots and Systems (IROS, 1988), International Symposium on Micro/Nano Mechatronics and human Science(MHS, 1989), IEEE Conference on Advanced robots and Social Impact(2005), System Integration International(2008), IEEE Conference on Cyborg and Bionic Systems (CBS, 2017), IEEE Conference on Intelligence and Safety of Robots (ISR, 2018). He has received many awards such as IEEE Robotics and Automation Pioneer Award (2004), IEEE Robotics and Automation Technical Field Award (2010), Medal of Honor on Purple Ribbon (2015), The Order of the Sacred Treasure, Gold Rays with Neck Ribbon (2022). IEEE Fellow (1995), SICE Fellow (1995), JSME Fellow (2002), RSJ Fellow (2004), VRSJ Fellow (2011), member of the Japan Academy of Engineering (2013).

Pig weights are important indicator for the healthcare and the economic operation of pig farms, and the development of a system to easily estimate these weights is desired. Although load cells are usually used for actual measurement in pig farms, it is not easy to guide pigs weighing more than 100 kg to the scales because many pigs do not like to get on the scales. Therefore, a convenient pig weight estimation system using RGB-D sensors has been developed. An RGB-D sensor is used as the sensing device for weight estimation. Weight estimation is performed on 3D point cloud data of photographed pig images. When capturing pigs, it is desirable to have a constant camera orientation toward the pigs However, it is not easy to always capture from the same direction because pigs move around quickly in the piggery. A method with a high degree of freedom in the capture direction by exploiting pig symmetry of the pig’s body is introduced in this paper. The system is applied for a wearing device using AR (Augmented Reality) glasses. In addition, its application to non-contact body measurements of low-birth-weight infants will be presented.

Kikuhito Kawasue Ph. D. is currently a professor at University of Miyazaki. Born in Nagasaki Prefecture in Japan. M.S., Nagasaki University (1989). Researcher assistant at Sasebo National College of Technology (1989). Visiting Scholar, Florida State University (1994). Ph.D. Nagasaki University (1996). Assistant Professor, University of Miyazaki (2000). Associate Professor, University of Miyazaki (2007); Professor, University of Miyazaki (2011). The "AI glasses that can see pigs' weight" was selected as one of the "Top 10 Agricultural Technology News in 2021" by the Ministry of Agriculture, Forestry and Fisheries of Japan. The technology of the "The AI glasses" has been translated into five languages and introduced under the name "Sow-ter" on the official website of the anime "Dragon Ball.

Large Language Models (LLMs) have recently gained considerable attention and popularity due to notable advancements and extensive media coverage, primarily driven by the success of commercial products. These models are extensively trained on large-scale corpora datasets, which have unveiled their unique and remarkable capabilities in processing and generating diverse media content, often exhibiting human-like capabilities. Thanks to their impressive results, they are now being integrated into many industrial pipelines, e.g., automated customer service systems, content creation tools, and advanced data retrieval and analysis platforms. However, LLMs are vulnerable to adversarial attacks, resulting in data leaks or system misuse for unintended purposes. Even recent models like ChatGPT, GPT-4, or Llama 2 remain susceptible and struggle to consistently prioritize initial developer and company guideline prompts. In this talk, I aim to provide an overview of the fundamentals of LLM and explore their various applications. The discussion will continue with analyzing the security risks associated with adversarial attacks and potential misuse. It will conclude with a broader analysis of concerns about the biases and hallucinations inherent in these models. The goal is to offer a comprehensive overview of the capabilities of large language models while highlighting the associated risks and challenges.

Antonio Emanuele Cinà has been an assistant professor (RTDA) at the University of Genoa, Italy, since June 2023. He received his Ph.D. (cum laude) in Computer Science from Ca' Foscari University of Venice in 2023, defending a thesis on the vulnerabilities and emerging risks arising from the malicious use of training data in AI. His research interests encompass all aspects of AI system security and the study of their trustworthiness, with primary expertise in training (poisoning) and inference-time (evasion) attacks. Recently, he has been investigating the capabilities of Generative AI models (LLMs), exploring the security aspects of these cutting-edge systems and how this technology can be integrated to optimize user applications.