Week #5: All Robots Break Loose

1. Implementation

This week, the team made progress in the various components that make up this solution. Firstly, the team made progress in the implementation of the multi-agent system, with the agents being able to coordinate between themselves in order to tackle the tasks that dynamically come up. Unfortunately, the competitive aspect could not be implemented this week, transitioning into the next one.

Moving on to the robotics, this week the team focused on setting up the restaurant environment, which at the moment appears as shown in screenshot below, and placing multiple TurtleBots inside it, with them being able to move around autonomously, without colliding. Next week, the team expects to be able to make the robots move from point A to point B on command, which will be crucial for the agents action to be executed.

Restaurant simulation environment

It must be noted that human beings are complicated to include in Gazebo and, following the professors' advice, a different TurtleBot model will be used to simulate the human clients moving around the environment, as the accurate simulation of humans is not a very important factor for this project.

Finally, regarding IoT, the team managed to configure an device with a pressure-sensitive sensor, capable of identifying when people are sitting down or standing up from a chair, which will come into play when the multi-agent system needs to plan how many agents are necessary at a given time in the restaurant. As it stands, the IoT devices being experimented by the team appear as follows.

IoT device experiments

2. Scientific Article

Having delivered the state of the art chapter last week, feedback was obtained from the professors this week, with the team improving the chapter based on that, while also writing the following chapter, where an overview of the solution is presented, all the way from the architecture and features to the benefits and disadvantages when compared with other available solutions.

3. Week retrospective

Yet another productive week comes to an end, with the progress made in the several fronts filling the team with positivity and motivation to move on with the project's development, as the upcoming weeks will surely be challenging ones.

Next week, the team expects to implement the competitive behaviour on the agents, implement point A to point B navigation on the robots, implement the necessary code for the IoT devices to communicate with a MQTT broker and write the Development chapter of the scientific article.

See you next week!