[agents] PhD positions on Human AI Teamwork at Delft University of Technology

[Myrthe Tielman]

For the TU Delft AI lab AI*Man on human-AI teamwork we are looking for 2 PhD candidates. The first on Fuzzy-Logic Based Mental Models in Human AI Teams, the second on Transparent and Traceable AI for Teamwork

The DAI Lab AI*Man is led by Anahita Jamshidnejad and Myrthe Tielman and is a collaboration between the faculties of aerospace engineering (AE) and electrical engineering, mathematics, and computer science (EEMCS) of the Delft University of Technology, the Netherlands. The main focus of the joint research is on human-AI teamwork.
The main challenges of human-AI teamwork include obtaining mutual understanding among the teammates and making optimal decisions to realize the team's global objectives. To achieve this, this multidisciplinary project will integrate various methods from control theory with knowledge-based methods from AI. Beyond this research goal, this project aims to build a bridge between the faculties of AE and EEMCS on the topic of AI, including both research and educational activities. Your home department will be the Interactive Intelligence department of EEMCS, the positions are for 5 years and involve direct collaboration with 3 other PhD students.


Fuzzy-Logic Based Mental Models in Human AI Teams position:
The research for this PhD centers on allowing AI team members to better understand their human team mates. In any team, mutual understanding is very important, but so is being able to explain what you know about the other. For this reason, the focus lies on knowledge-based AI techniques, and in particular using fuzzy logic. Your research on this project will explore how we can create new reasoning frameworks to understand human decision making in team contexts. Some examples of the types of tasks involved are: studying human-decision making within the realm of human-robot teams, translating human choice behavior in this realm to fuzzy logic models, adapting a robot’s behavior based on these fuzzy models and updating these models based on feedback from interactions.

Transparent and Traceable AI for Teamwork:
The research focus for this PhD is about making AI systems (such as robots) transparent, traceable and controllable to their human teammates. Lack of understanding of what teammates are doing impairs the functioning of a team. For AI, that creates the challenge of explaining complex decisions to humans, but also sets the goal of making these decisions traceable and controllable in case of mistakes made by the AI. The focus will lie on knowledge-based AI for instance using formal reasoning models, as these allow for higher levels of transparency than, for instance, machine-learning based models. Examples of the types of tasks involved are: explaining decisions in human-robot teaming, updating AI models based on feedback from interactions, transforming decision traces to transparent explanations and, when/how to allow explicit human control over AI teammates.

Requirements:
Completed Master’s degree in artificial intelligence, logic or a related field;
An affinity with logic, knowledge-based AI and human-robot teamwork related research;
An Independent, results driven, motivated personality with the ability to work in an interdisciplinary team;
Proficiency in written and verbal English;
Good presentation skills;
An affinity with teaching and guiding students.


For more information, contact details and to apply, see the full vacancy descriptions at:
https://www.academictransfer.com/en/292380/phd-position-fuzzy-logic-based-mental-models-in-human-ai-teams/
https://www.academictransfer.com/en/292379/phd-position-transparent-and-traceable-ai-for-teamwork/




Dr. Myrthe Tielman
Assistant Professor
m.l.tielman at tudelft.nl

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cs.umbc.edu/pipermail/agents/attachments/20200619/1f2c4841/attachment.html>