TL;DR:
- RiTA 2023 Conference discusses self-driving car safety and Machine Learning (ML) algorithm vulnerabilities.
- Professor Xiaowei Huang presents a co-simulation method to enhance the ML algorithm’s ability to identify unsafe conditions.
- Over 100 global participants gather at XJTLU Entrepreneur College for knowledge exchange.
- Dr. Youjia Wu highlights RiTA’s role in fostering research collaborations.
- Professor Eng Hwa Yap emphasizes XJTLU’s role in promoting innovation and entrepreneurship.
- RiTA, an annual event, continues to bridge academia and industry in the field of robot intelligence.
Main AI News:
The 11th International Conference on Robot Intelligence Technology and Applications (RiTA 2023) unfolded from the 6th to the 8th of December at the prestigious Xi’an Jiaotong-Liverpool University’s Entrepreneur College (Taicang). The conference, co-organized by XJTLU, the University of Liverpool, and the Korea Advanced Institute of Science and Technology (KAIST), revolved around the theme “Robotics and AI for Digital Futures.”
One of the pivotal discussions at the conference was centered on enhancing the safety of self-driving cars. Renowned expert, Professor Xiaowei Huang from the University of Liverpool, delivered a keynote speech, where he delved into his groundbreaking research dedicated to ensuring the safety of autonomous vehicles. Professor Huang’s research unveiled potential vulnerabilities within the Machine Learning (ML) algorithms employed by self-driving cars, particularly in their interpretation of unfamiliar environments. These weaknesses might result in the system failing to recognize hazardous conditions.
In his address, Professor Huang articulated the issue, stating, “The ML algorithm may need to work on many environments that are different from the environment where the training data was collected.” To tackle this challenge, Professor Huang introduced a novel approach known as the co-simulation method, aimed at augmenting the ML algorithm’s capacity to detect unsafe elements, even in uncharted territories.
This innovative approach involves two parallel systems: one employs pure simulation for ML algorithm training, while the other gathers real-world data to continually update the algorithm. The synergy between these components enables the accumulation of compelling evidence, ultimately reducing the presence of unsafe factors when autonomous cars are in operation.
RiTA 2023 gathered more than 100 academics, industry experts, and students from across the globe, including China, Germany, Malaysia, and South Korea, all converging at XJTLU Entrepreneur College (Taicang). The conference attracted over 60 papers submitted by a minimum of 20 institutions worldwide, fostering an environment where academic and industry experts could exchange insights.
Dr. Youjia Wu, an Applied AI Scientist at the Bosch Center for Artificial Intelligence and one of the conference’s keynote speakers, expressed the significance of RiTA in fostering communication and potential research collaborations with academic communities and fellow industry professionals.
Professor Eng Hwa Yap, Dean of the School of Robotics and the School of Intelligent Manufacturing Ecosystem, emphasized how XJTLU serves as a catalyst for global dialogue and innovation by bringing together international experts, researchers, and academics. This exchange of ideas not only enhances the institution’s reputation but also nurtures a mindset of innovation and entrepreneurship among students.
International conferences like RiTA have proven to be invaluable resources for students. By engaging with leading experts, students gain firsthand exposure to the practical applications of entrepreneurial thinking in cutting-edge fields. Interactive sessions and workshops provide students with the unique opportunity to network with industry leaders, further fueling their passion for innovation and entrepreneurship.
Since its inception in 2012, the Korea Advanced Institute of Science and Technology (KAIST) has been organizing RiTA annually. Over the years, RiTA has evolved into a pivotal event that brings together scholars and industry experts in the realm of robot intelligence technologies. This platform not only facilitates knowledge exchange but also serves as a fertile ground for nurturing collaborations that advance the field of robot intelligence.
“As the field of robotics is inherently multidisciplinary,” notes Professor Han-Lim Choi from KAIST, “RiTA distinguishes itself by anchoring firmly in robot intelligence and expanding its horizons to encompass the broader spectrum of robot systems.” This commitment to pushing the boundaries of robot intelligence is what continues to make RiTA a cornerstone event in the world of robotics and AI.
Conclusion:
The RiTA 2023 Conference showcased significant strides in enhancing self-driving car safety through innovative ML algorithm approaches. This development highlights the increasing importance of robust AI technologies in the automotive industry, signaling a growing market demand for AI-driven solutions to ensure the safety and reliability of autonomous vehicles. The collaborative spirit of the conference also underscores the potential for future research partnerships, further driving advancements in the field.
