The "handling qualities", or ease of control, of human controlled vehicles are difficult to objectively characterize. The subjective nature of human perception to changes in vehicle dynamics confounds the certainty in hypothesized relationships between plant dynamics and the handling qualities. For example, humans can even be tricked into incorrect judgements when there is nuance in varied vehicles dynamics. Additionally, perception can be affected differently for different tasks and for equivalent order plants with large differences in the dynamics. A first step to reaching relationships between plant dynamics, task type, and perception of handling is to understand what the threshold precision of subjectively reported perception of a human subject under controlled situations. This has been explored for laboratory tasks, such as tracking a single degree of freedom motion on a screen with a joystick [Wei2020] and aircraft stall [Smets2019], but not yet in more complex experimental scenarios.

The bicycle-rider system is well suited to attempt perception threshold identification due to the ability of constraining the plant as a SIMO system and that it is a "lab scaled" and a low cost experimental platform. The bicycle provides unique plant dynamics that can be exploited, such as variable stability, non-minimum phase behavior, and non-trivial system order. Some prior work with a bicycle has been performed in [Kresie2017]. There are two primary options for varying the dynamics of the vehicle: 1) manually change specific physical aspects of the vehicle or 2) to make use of the Bicycle Lab's steer-by-wire bicycle in which the open loop dynamics between the handlebars and fork can be implemented in software. Both have been done in the lab before and there are merits to each. A method of setting precise and repeatable plant dynamics will be required.

The primary goal of the project will be to develop experimentally derived measures of the perception thresholds to changes in vehicle dynamics, i.e. what's the smallest change in a physical characteristics that riders can successfully detect.

Experiments will have to be carefully designed such that the rider does not make perception judgements based on preconceived ideas about how physical characteristics relate to handling and appropriate control tasks (maneuvers) will have to be chosen that maximize the ability to make objective measures of control success. The dynamics will have to be able to varied precisely for repeatability and also in a way that both randomizes the order of changes and hones in on threshold limits in with a minimal number of trials. Multiple riders will need to be evaluated to increase certainty in the perception measures. Riders will need to be coaxed into utilizing similar passive biomechanics or these differences will need to be measured. Lastly, careful design of extracting the subjective assessment from the subject for a given trial will need to be implemented.

We will collaborate with Rene van Paassen in Human Machine Systems in Aerospace Engineering.

How To Apply

If you would like to apply for this project, please send an approximately half page letter explaining your motivations and interest in the lab and project, CV or resume, a list of courses you've taken, the name of your MSc track, and any other relevant information to j.k.moore@tudelft.nl.