As part of my PhD, I conducted a proof-of-concept study. The purpose of the study was to assess the performance of a robot that integrated the deadband-sampling event-based processing method with the custom electronic nose (e-nose) and the one-dimensional Braitenberg-style control algorithm. Confronting live data in a real-world environment, facing noise-ridden, turbulent-induced signals tested the success of this approach.
This post is a part of the supplementary material for my thesis. If you are interested in seeing more details, I will soon provide a reference to a published preprint.
In the meantime, details on the processing method and the custom e-nose can be found on this post.
Braitenberg-style Control Algorithm
In summary, the algorithm defines how the speed of the motors is adjusted based on its signal processing state, specifically whether it is experiencing an “ON” or “OFF” event:
- “ON” Event: The speed is increased but capped at a maximum limit.
- “OFF” Event: The speed is reduced but not allowed to drop below a certain threshold.
- No Event: If neither event is occurring, the speed remains constant.
As a result of this algorithm, the robot operates in one-dimensional space, inspired by the Braitenberg vehicle thought experiment. In this style of navigation, the robot moves towards detected stimuli, reacting only to locally sourced olfactory signals.