Irfan Refai
Student Projects
For a complete list of open student projects, please refer to our department website.
Learning human-exosuit interaction forces

Learning human-exosuit interaction forces
open

Soft wearable robots, or exosuits, are more comfortable and wearable. However, they are tricky to model. Your challenge, if you so choose, is to develop machine-learning methods to improve the modelling of the human-exosuit interface.
Physico-digital twins for Human-Exosuit Interaction

Physico-digital twins for Human-Exosuit Interaction
open

Can a combined physical and digital twin of your limb help us in designing better exosuits? How close to reality can we get in terms of physical and digital modelling?
Optimizing design of soft exosuits (for space travel)

Optimizing design of soft exosuits (for space travel)
open

Soft wearable robots have the freedom to wrap around the contour of the user. But, which design optimizes comfort? and which one optimizes biomechanics? and which is suitable for space travel? You will be tasked with finding this out!
Biomechanical Modelling for Edge Devices

Biomechanical Modelling for Edge Devices
open

Wearable movement sensors collect a plethora of data from users using different modalities to develop digital twins. The models required to personalize these digital twins are computationally heavy. Federated learning can be a potential solution in reducing their complexity for edge devices. Your mission is to develop techniques that optimizes deployment of large-scale continual models on edge devices.
Biomechanical Modelling from Multimodal Sparse Sensors

Biomechanical Modelling from Multimodal Sparse Sensors
open

Imagine you are drifting in space and require a precise orientation to get from point A to point B in your space ship. There are several sensors sparsely spread out. What's the ideal way to fuse the information from them?