Young Scientist: Machine vision to aid self-driving cars

Teen builds robot from off-the-shelf devices and writes software needed to link them

 

A Dublin student has built technology that provides a form of machine vision and could serve as the control mechanism for self-driving cars.

Alec O’Brian is a 16-year-old fifth year student at Castleknock College and he developed a self-guiding robot that uses computer vision to move without human control through the environment.

“It has spatial awareness,” said Alec, who assembled a number of off the shelf devices and then wrote the software needed to link them.

It took three or four months to put the six-wheeled robotic vehicle together and it has a range of facilities that allow it to move autonomously.

Stereo vision

It has stereo vision and also has ultrasonic range finding in order to gauge distances to objects and also the size of objects it encounters.

This allows the robot to move around objects and choose a path that won’t get it stuck in narrow spaces.

Its vision system allows it to track objects as they move. It also has a magnetometer so that it knows its direction of travel. It will soon be fitted with GPS so it can use this system to move to locations or navigate in a given direction.

It could easily be connected to a general robot, said Alec. “They are machines, they are blind, but this gives vision to any machine.”

He plans to integrate some of these systems into a drone, allowing it to fly through an environment while avoiding objects. The system would also have ready application in cars that can drive themselves, he said.

Two students from Synge Street CBS in Dublin have pursued a very different kind of project, coming up with a way to monitor and hopefully control the spread of a very damaging infection that affects coffee plants.

Natural control

The goal was to come up with a natural control method and rather than chemicals and sprays it depends on mathematics.

Gabriel Barat and Adrian Wolniak are sixth year students and they developed a mathematical model for the spread of a fungal disease called coffee rust. “It is a very serious problem and is endemic in Central America,” said Adrian.

However the disease is complicated because it is linked to other other diseases and to an ant.

An existing model includes two of these, coffee rust and white halo fungus, and this is used to track disease and predict spread.

The students went further, building a model that includes these but also two more vectors, green coffee scale and the Azteca instabilis ant that uses the scale for food.

Powerful model

This has given them a much more powerful model for understanding the movement of the rust and its impact on plantations, the students say.

The exhibition reaches its high point on Friday evening when the judges deliver their verdicts and the BT Young Scientist 2016 is announced.

The Minister for Education and Skills Jan O’Sullivan will be at the RDS to hand over the top prize for the best individual or group, a cheque for €5,000 and the BT Young Scientist perpetual trophy. They will also have the opportunity to represent Ireland in the EU Competition for Young Scientists taking place this autumn in Brussels.

The next placed best individual or group will receive a cheque for €2,400 and a BT perpetual trophy. Runner-up prizes are €1,200.

The exhibition is open to the public on Friday and Saturday with tickets at the door costing €6 for students and concessions, €12 for adults and €25 for an adult ticket.