University logo SPS / BIASlab
  • Home
  • Research
    • Projects
    • Publications
    • Blog
  • Teaching
    • 5SSD0 (2024/2025)
  • Software
    • RxInfer.jl
    • ForneyLab.jl
    • GraphPPL.jl
    • ReactiveMP.jl
    • Rocket.jl
  • Team
  • Join us
    • Starter's guide
    • Open projects
  • Contact

Tim Nisslbeck

PhD candidate, TU Eindhoven

Interests

  • Machine Learning
  • Statistical Inference
  • Neuroscience
  • Robotics

Education

  • MSc in Computer Science, 2018

    Friedrich-Alexander University Erlangen-Nuremberg

  • BSc in Computer Science, 2016

    Friedrich-Alexander University Erlangen-Nuremberg

Biography

My general long-term research goal is to develop fully autonomous agents that can solve real-world problems. Specifically, I aim to build agents that can control robots to automatically perform tasks, thus reducing the workload that was previously handled by humans. In order for the agent to become fully autonomous, it must be able to make decisions in real-time. Often, such decisions must be made under uncertainty due to numerous unknown factors in its dynamic environment. In high-risk applications such as robotics, the agent’s decision-making process must also be safe and explainable.

Currently, my work as a PhD student at TU/e focuses on designing the artificial intelligence behind these agents, which is a multi-disciplinary endeavor. My research involves combining knowledge from (Bayesian) machine learning with theories from neuroscience to develop such intelligence. The resulting agents should then be able to learn and act similarly to humans, while their decision-making is both safe and explainable. To demonstrate the capabilities of these agents, their methods will be implemented on a quadrepedal walking robot as part of the FEP Quad project.

Previously, I worked as a research scientist at Fraunhofer on improving automated driver assistance systems for autonomous vehicles by employing deep reinforcement learning methods.

Publications

Coupled autoregressive active inference agents for control of multi-joint dynamical systems

Tim Nisslbeck, Wouter Kouw
IWAI 2024
September, 2024
Details PDF Code

© BIASlab, 2024 · Partially powered by the Academic theme for Hugo.