- Published: 07 January 2022
The University of Macau (UM), to help develop Macao into a smart city and also in the diversification of the local economy with technological advancements, is creating numerous novel applications on the basis of artificial intelligence technology. This includes unmanned marine vessels, self-driving buses, image forgery detection, industrial robots, and software for smart tourism.
One of these projects is the self-driving project shouldered by the State Key Laboratory of Internet of Things for Smart City (University of Macau) and its associates from Macao and mainland China. The scientists were led by Professor Xu Chengzhong, dean of UM’s Faculty of Science and Technology.
The researchers launched Macao’s first‑ever self‑driving bus as a testing vehicle at the university to self‑driving cars safer. Professor Xu’s group observed encouraging results. The team has published numerous articles on related technologies in leading academic journals. They have also presented some other papers at top international conferences.
The UM scientists have put forth a novel algorithm known as Re‑Initialising the Fully‑connected LayEr (RIFLE) to teach cars to make smart decisions. Also, self‑driving cars should be capable of identifying everything around them.
Convolutional neural networks generally carry out these tasks efficiently, however they are unsafe when confronting unclear objects or cyberattacks. Hence, Professor Xu’s group created the “LAFEAT” algorithm, which makes CNNs more powerful to adversarial noise or attacks.
A group of researchers headed by Carlos Silvestre, a professor in the UM Department of Electrical and Computer Engineering, created the next‑generation marine vessels that are capable of performing tasks on their own. The scientists of the group have also generated algorithms that enable unmanned surface vessels (USVs) to return to their starting point when running out of power, evade obstacles, and handle unexpected situations.
The team won the first prize at an international intelligent vessel competition in 2020, and since then the group was involved in designing underwater robots with sonar sensors that can create high‑resolution underwater images even in low‑visibility conditions. During typhoons and other natural disasters, this vessel can be used to check if there is any damage to the underwater facilities in the waters of Macao where visibility is less.
Xu Qingsong, a professor in the Department of Electromechanical Engineering, heads the group to develop innovative intelligent robots. Professor Xu earlier created a second-generation disinfection robot known as “Smart Cleaner”. Professor Xu states that UM has filed a patent application for the robot technology. The university is also communicating with companies for patent transfer, to introduce the robots on the market.
The researchers are also creating a next‑generation industrial robot with three‑dimensional visual perception and compliant force control. The group intends to use the novel robot to polish cars, assemble parts of devices, and maintain aircraft. Professor Xu’s group has also created an intelligent micromanipulation robot system that spontaneously controls microgrippers and microinjectors at the micro- to nano-scale for cell operations like gene editing.
The research is also aided by the Zhuhai UM Science & Technology Research Institute to convert the research results into marketable products.
Associate Professor Zhou Jiantao from the Department of Computer and Information Science, who is also the interim head of UM’s Centre for Artificial Intelligence and Robotics, created an advanced algorithm, with the help of the Super Intelligent Computing Centre at UM, to identify image forgery.
Overtaking more than 1,500 rivals at an international competition, the scientists continued to enhance the algorithm under a one‑year project with the e‑commerce giant Alibaba. The tech company sponsors Professor Zhou’s group to generate an algorithm that precisely identifies forged parts of images, even when resized, compressed by various media, filtered, or consisting of added noise.
Professor Gong Zhiguo, head of UM’s Department of Computer and Information Science, and his postgraduate students designed novel algorithms that utilize social media data to analyze the sentiments of the tourists to recognize their favorite places and activities in Macao, along with identifying and predicting emergent events.
Moreover, the scientists utilized information from social media to evaluate the density of tourists in various areas over the course of the day. According to Professor Gong, such statistics could inform government planning for tourist facilities and the transport system.
Professor Gong’s team intends to launch another mobile application that reveals real‑time congestion levels on a map and makes predictions to assist tourists and residents to get around the city and also avoid traffic jams—aiming to provide finer and smarter travel experiences for tourists in Macao.
Autor(en)/Author(s): Aimee Molineux
Quelle/Source: AZO Robotics, 30.12.2021