Today, people can turn on the lights, order groceries, and change the channel on the television using only their voices. Cars can automatically break and slow down to adjust to traffic flow. Companies are starting to utilize data and innovative technology to transport patient care in the healthcare industry. This is being done thanks to researchers such as Chenglin Miao, assistant professor of Computer Science, who is working on building secure connections between humans and the physical world using artificial intelligence (AI) techniques and the Internet of Things (IoT).
IoT is a new networking model that connects humans and the physical world using ubiquitous sensing, computing, and communicating devices. With IoT having various civilian and military applications, countries and companies worldwide are utilizing research to transform industries and daily life. The result is that the devices collect a large amount of data and make smart decisions.
Researchers have started developing various intelligent Internet of Things systems using IoT devices, such as autonomous driving systems using artificial intelligence techniques. However, with a large amount of data being collected come security challenges. This is the focus of Miao’s research: how can we find the security vulnerabilities of AI techniques when applied to real-world IoT systems, and how can we address them? What countermeasures can we utilize?
For Miao, the crust of his research is to develop practical and easy-to-use attack methods for analyzing the security vulnerabilities of AI-enabled IoT systems. The better systems we have to analyze the attack methods, the better tools we can use to develop effective countermeasures against malicious attacks.
This research has its challenges. The performance of AI-enabled systems can easily be affected by many uncertain factors, such as environmental noise and interference from surrounding objects. Many AI models can easily be fooled by attackers in the physical world. With autonomous driving systems, Miao’s recent work has demonstrated that attackers can easily fool the autonomous vehicle relying on AI techniques by slightly changing the driving environment, such as placing simple objects such as advertising boards at a few specific locations in the physical space. The result will be the autonomous system failing to see a car in front of it, thus creating a dangerous environment for the passengers.
Miao hopes that his research can be used as a key enabler for developing and deploying safe IoT systems for future human life. Additionally, he believes his research findings may benefit other research areas related to AI, such as robotics and human-computer interaction. Through research such as this, we can become a safer, more protected technological society in many growing aspects of our lives, from work to homes.
Chenglin Miao is an assistant professor of computer science at Iowa State University. He received his Ph.D. in computer science and engineering from the State University of New York at Buffalo in 2020. His research mainly focuses on the security and privacy aspects of Internet of Things (IoT) and artificial intelligence (AI). His research work has been published in various top venues such as CCS, MobiCom, SenSys, MobiSys, NeurIPS, AAAI, and IJCAI.