The Eindhoven Artificial Intelligence Systems Institute (EAISI) is funding 2 PhD positions for the highly innovative
BayesBrain project, in collaboration with the departments of
- Mechanical Engineering,
- Biomedical Engineering,
- Electrical Engineering, and
- Mathematics and Computer Science.
The challenging goal of the project is to design an in-vitro AI computer which leverages the computation of neural cultures to solve real-world AI problems. This hybrid approach will consist of an in-silico Bayesian control agent and a Brain-on-Chip device, which shall communicate via the common principle of Free Energy Minimization (FEM).
The announced PhD position focuses on devising the hard-ware of a
compartmentalized microfluidic chip on microelectrode arrays (MEAs) that form the
interface to the in-silico control agents for the purpose of building a hybrid AI computer. For this position, relevant background knowledge and hands-on skills on
neural cell culture, integrated microfluidics, sensors & actuators (MEA technology), and
fast and high-quality data acquisition systems is required. The successful candidate will critically review the state of the art (high-density versus low-density MEA tech) and creatively design a simple-to-use microfluidic platform that will have a central role in building the hard-ware of the hybrid AI system of connected neural circuits, which we individually can address in the system by voltage or current injection. Furthermore, the techniques developed will have a wider applicability in
material and toxicity screening,
drug development and
personalized medicine and the
discovery of novel neuro-therapeutic interventions. The successful candidate will be hosted at the Department of Mechanical Engineering and co-supervised by the Department of Biomedical Engineering.
Paired with the announced position is a second PhD position hosted at the Electrical Engineering department and co-supervised by the Department of Mathematics and Computer Science. The second PhD candidate will focus on the development of in-silico Bayesian control agents via probabilistic programming. The two PhD students will collaborate closely to achieve the final goal of developing a hybrid neural/in-silico AI computer. In particular, the two PhD students will work together on the development of interfaces between the in-silico Bayesian control agent and the Brain-on-Chip device.
The preferred starting date of this position is February 2022.