Mechanical Engineering (sensing the world with sound) Taster Session

In the pitch black, a bat speeds through the air and swoops around a tree to eat a tasty insect snack. Bats avoid trees and find snacks just by using sound. Scientists have been inspired by bats to use sound to detect flaws in the ancient UK sewage system. We are now designing a swarm of small robots that use sound to detect cracks and blocks in sewers. In this talk, you will learn a physicist's approach to use sound, like a bat, to find things in the dark. Will you be able to recognise the sound of a blocked sewer?

The session will cover 1) a fun application, the UK sewers, 2) I will teach how to approach problems like a physicist, and 3) there will be lots of videos and some interaction with the audience.

1.) Maintaining the UK sewer system is a huge issue. Every time we suspect there is a major leak we need to dig up large parts of our roads, causing traffic jams, air pollution from dust, and the whole operation is very costly. To make matters worse, much of our sewers are ancient, falling apart, and we have no idea what pipes are where. Not to mention that it is difficult to convince people to go down there to inspect the sewers. How do we now maintain this ancient labyrinth? A team, led by the University of Sheffield, have the answer: we will develop a swarm of small robots to swim through our sewers and detect cracks and blocks.

2.) After presenting the application, I will pose the question: how do these little robots detect things in the dark? I will then show how, with a physics approach, we can break this problem down into simple manageable smaller problems. These small problems involve simple trigonometry and the solution is illustrated with lots of videos from simulations. From there, we build back up to the complex problem.

3.) The talk will involve lots of videos and have strong visual aids throughout. Videos will include bats flying around, sound sensors in pipes, and computer simulations. I will ask the audience to hear sounds and guess what they heard. Did the sound come from a blocked sewer or not? To finish we end with how machine learning is better at recognising sounds then we are.

Approx itinerary:
4.00pm - 4.15pm: Log in and get comfortable
4.15pm - 4.30pm: Welcome and introduction to the department
4.30pm - 4.55pm: Sewer talk. Students vote on which sounds correspond to a block sewer.
4.55pm - 5.05pm:Interactive “draw the lines of pressure exercise”
5.05pm-5.25pm: Q&A with Mechanical Engineering students
5.25pm - 5.30pm: Closing remarks and next steps