University of Southampton Eurobot 2020 team project update
Who are we and What is Eurobot?
We are a team of four third-year mechanical engineering students from the University of Southampton. After performing well last year in the autonomous pick and place robotics competition, Eurobot, we decided to form a team and partner with RS to surpass last year’s achievements.
The 2020 theme is ‘Sail the World’, where teams accumulate points by having dynamic mechatronic systems retrieving and delivering coloured cups, called buoys. In addition, a stationary mechatronic system, that must be activated by the autonomous robots, representing a lighthouse, can be developed for additional points.
You can find our first project update with more details on the competition and our background here.
How is the project progressing?
Primary Robot - Rose
As a reminder, we have a larger primary robot named ‘Rose’, which was devised to pick up a total of four grounded buoys and subsequently place those of the appropriate colour on the correct fairway.
The main challenges in the development of Rose have been the wiring, the functioning of its electronic components, and the fine-tuning of the mechanisms.
Rose’s wiring has been completely redone since the prototype stage. Clean and organised wiring is critical in mechatronics projects to enable us to spot problems as fast as possible when they occur. Having neat wiring also helps any additions of electronic components to the circuits or any modifications we wish to do in the future.
A challenge the team faced was the utilisation of a large stepper motor. This is a component none of us had any experience with, and the large current demand posed several problems. The main issue was the compatibility of the stepper with stepper drivers. At high loads, some stepper drivers were unable to meet the current demand of the stepper, sometimes even blowing the circuit. The use of the RS A4988 stepper driver allowed us to overcome this issue, and run the stepper in a smooth and controlled way as demonstrated in the video below.
The team ran into further challenges when testing the mechanisms. Some of these had friction problems, affecting the speed and effectiveness of the mechanisms. The use of igus slip tape from RS solved this issue. Another problem occurred when trying to activate the windsocks. This was resolved by travelling at a slower speed during activation.
Secondary Robot - Jack
Our other robot, ‘Jack’, has two compartments and guiding arms which could store a total of four buoys, where the compartments can be utilised to separate the buoys by colour.
Since Jack was mostly finished by the last time we wrote the article, there has been little development of the physical system. Instead, the team has created the path for the robot and can score a large number of points. A video has been linked below of a test run.
Stationary Robot - Lighthouse
The lighthouse is another aspect of the competition which scores a significant quantity of points at 15. It involves a rotating light rising from an initial height of 300 mm to a final height of 700 mm. The team realised that a “seesaw” mechanism would be able to fulfil the task. Our initial 3D CAD design is shown below. A simple push-button, in the middle, which uses a microswitch is the method of activation. It was designed this way to give a consistent result.
When choosing the mechatronics to provide motion, the team chose effective components which were unlikely to fail. The motor chosen was more than adequate for the task and to remove any uncertainty with the finishing point of the arm, a microswitch was added. This allowed for the motor to be stopped the instant the arm was in the correct final position. Using this method as opposed to activating the motor on for a set amount of time allowed for the greatest possible consistency. With this final design, the team is confident that we would be able to score 15 points on every run. A video of the lighthouse in action can be seen below.
Computer Vision System - Obstacle Avoidance
As mentioned in the last article, a significant difference in strategy this year is the plan to be the first UK team to utilise computer vision in the competition. The rulebook states that QR codes must be present at the top of each robot, so a vision system is being developed to capitalise on this rule.
The most significant difficulty we encountered when developing the vision system was not actually in the detection and recognition itself, but the underlying communication system. We originally planned to use Arduino Bluetooth modules but quickly realised that there was a considerable latency in transmissions. To resolve this problem, we used some of the Raspberry Pi Zeros supplied by RS to handle communications. The result of this change is a system that is robust and highly responsive to obstacles. A video of a successful system test can be seen below.
What happens next?
Unfortunately, the abrupt closing of university facilities due to COVID-19 has postponed any further work on the robots. This has been quite disappointing as there was less than a week’s worth of work left before the robots are competition ready. All that is left to do is to optimise the paths of each robot.
Although the world final is still due to go-ahead in La Roche sur-Yon, France this summer, the state of the UK final is still uncertain. There is even a possibility that Middlesex University will take online entries. Whatever the method, the team is eager to demonstrate the effectiveness of the robots and succeed in this year’s competition.