Research

Research Topics

The lab undertakes research in a broad range of disciplines, aiming to integrate multiple subjects of studies with advanced machine learning methods. Currently, the lab is focusing on the following projects.

1. Odour Sensing

Olfactory information is an important ability for living creatures. Animals can utilise it to localise foods, water and mates. However, the slow reaction speed of conventional gas sensors is a critical issue for an odour tracking robot. The electroantennogram (EAG) is a technique that can extract neural signals from an antenna of an insect for detecting odour plumes. However, due to its small amplitude, the EAG signal can be easily influenced by the surroundings. To obtain accurate detection results, special algorithms were developed to filter the signal recorded in noisy environments. This technique can hence be applied to flying robots for odour tracking later.

Keywords: odour sensing, electroantennogram, insect antenna

2. Analysis of Flapping Motion

Understanding the causal impacts among various parameters is essential for designing flapping motion systems, such as micro aerial vehicles (MAVs) and autonomous underwater vehicles (AUVs). Computational fluid dynamics (CFD) simulation enables us to analyse and compare the flow fields under different motion patterns. However, obtaining precise results often requires substantial computational time. To accelerate the analysis process, machine learning methods were introduced to reduce computational costs. Moreover, optimisation algorithms were implemented to identify effective motion strategies without the need for conducting thousands of simulations.

Keywords: flapping motion, biomimetic propulsion, machine learning, optimisation algorithm