Bell vs yell: comparative ease of localisation of acoustic warning strategies for bicycles

As with many courses, my course has a long-form dissertation project in third year. I chose to do mine on the ease of localisation of voices and bicycle bells, with application to not hitting pedestrians while cycling. To make a long story short, it turns out that the human voice is much easier to localise than a bell, but people make localisation judgements maginally faster for certain types of bell.

The project involved a lot of research, planning, programming, testing, changing of plans, and general head scratching, but it was honestly one of the best thigs I did at university. Working independently on a large project for most of a year taught me a lot of things about a lot of things, not least time and information management. It was hard at times, but nothing beats the moment of seeing some resullts fall out of your very first basic analysis pass and realising "aha, it worked".

You can view and download the PDF here.