"I enjoy working in the welcoming and supportive atmosphere that SFU provides. Apart from that, Burnaby Mountain (where my wife and I also live) provides for some stunning scenery."

Meet More Students in Applied Sciences

SEE MORE PROFILES

Curate your digital footprint

Want to be featured on our website? Complete our online submission form.

Submit your profile

Postdoctoral Profile: Jens Classen

December 10, 2021
Print

School of Computing Science Postdoctoral Fellow in the Faculty of Applied Sciences

I am a postdoctoral researcher at the School of Computing Science at SFU under the supervision of Prof. James Delgrande. My research is in Knowledge Representation and Reasoning, a subarea of Artificial Intelligence that studies how an agent's knowledge can be represented symbolically and subsequently manipulated through reasoning algorithms. The focus of my work has been on reasoning about action and change, reasoning about beliefs, planning, agent program verification, and (more recently) machine ethics. In addition, I am a sessional instructor at SFU, and have taught courses on Programming as well as Artificial Intelligence. Before joining SFU, I completed my PhD and worked as a postdoctoral researcher under the supervision of Prof. Gerhard Lakemeyer at RWTH Aachen University, Germany. Moreover, I worked as a software developer for mobility web applications and a visiting lecturer for courses on Knowledge Representation.

WHY DID YOU CHOOSE TO COME TO SFU?

When I saw that my current supervisor, Prof. James Delgrande, was looking for a postdoc to work with him on a project for combining the areas of belief change and reasoning about actions, I jumped at the opportunity to apply for the position, and luckily was accepted. He is one of the leading researchers on belief change, and I had a strong background in reasoning about action, so it seemed like a natural fit. The research we are doing addresses a reoccurring issue one encounters in many practical applications: Agents have to be able to deal with situations where their actions fail, where their sensors report incorrect or conflicting information, and where they fail to fulfill certain tasks or obligations due to unforeseen circumstances.

HOW WOULD YOU DESCRIBE YOUR RESEARCH OR YOUR PROGRAM TO A FAMILY MEMBER?

When building intelligent artificial agents such as mobile robots, symbolic representations have the advantage that they constitute a middle ground between the formal, mathematical way in which computers function, and the imprecise, yet very expressive natural language representations we humans use in our everyday lives. Having symbols (words) with a well-defined, mutually understood meaning for objects (e.g., a cup of coffee), their relations (e.g., the cup being on the table), and the actions that can be applied to them (e.g., picking up the cup from the table) makes it easier for humans and robots to comprehend one another. Compared to non-symbolic approaches such as neural networks, which are often considered to be "black boxes", this in particular allows more directly for explainability: For example, when being asked why it chose to pick up the cup from the table, a robot will be able to explain that it is because it intends to bring the cup over to you in order to serve you coffee. In my research, I have studied and developed different forms of such symbolic representations (in mathematical terminology: logics) that allow agents to reason about various aspects of themselves and their environment. This includes questions such as what an agent knows about the state of its environment, which actions it can do, how these actions affect the environment, how actions and sensing affect the agent's knowledge of the environment, how to come up with a course of action to accomplish a certain goal, whether a complex plan satisfies certain constraints such as safety or liveness, and more. During my postdoctoral work at SFU, I have specifically looked at possibilities for including ethical considerations, addressing questions such as what the agent is allowed, forbidden, or obligated to do, as well as ways of dealing with uncertainty, failing actions and sensors, and conflicting information.

WHAT ARE YOU PARTICULARLY ENJOYING ABOUT YOUR STUDIES/RESEARCH AT SFU?

I enjoy working in the welcoming and supportive atmosphere that SFU provides. Apart from that, Burnaby Mountain (where my wife and I also live) provides for some stunning scenery.

HAVE YOU BEEN THE RECIPIENT OF ANY MAJOR OR DONOR-FUNDED AWARDS? IF SO, PLEASE TELL US WHICH ONES AND A LITTLE ABOUT HOW THE AWARDS HAVE IMPACTED YOUR STUDIES AND/OR RESEARCH.

I was awarded the Borchers Badge at RWTH Aachen University for completing my doctoral degree with distinction. I am very grateful for the recognition, and it strengthens my motivation for pursuing an academic career.