Coffee shop bot serves up a win for Leeds team at SciRoc Challenge

A team from the School of Computing has taken home a prize at the inaugural Smart Cities Robotics Challenges (SciRoc) in Milton Keynes.

The team, led by Dr Matteo Leonetti and coordinated by PhD student Logan Dunbar, achieved the highest score in one of the event’s challenges, which tasked them with programming a robot to deliver orders at a café. The team also included undergraduate students Mohammed Alshamassi, Jit Hong Cheah,  Georgy Gunkin, and Joe Jeffcock.

SciRoc is an EU-H2020-funded project supporting the European Robotics League, particularly ERL Smart Cities. Contestants are invited to explore and solve key challenges of integrating robotics into smart city infrastructure.

The University of Leeds was represented by the Leeds Autonomous Service Robots (LASR) team along with its robot TIAGo, overseen by the Sensible Robots research group. This group’s research focuses on intelligent decision-making and machine learning for robots.

For this challenge, the team had to program a robot capable of autonomously inspecting the tables at a café and determining which ones were free, served or waiting to order. It also had to take orders from customers and direct new customers to empty tables.

The robot’s modelling behaviour incorporated Petri Net Plans, a framework for programming artificial intelligences. Google’s Dialogflow software was used to allow the robot to greet and interact with customers, while YOLO was used for object recognition.

The SciRoc event marked TIAGo’s first adventure outside the lab, and the LASR team were extremely pleased to be awarded top marks by the judges and the prize for winning the challenge.