Autonomous robot for indoor climate monitoring and cleaning
Project ID: AEE-2021
Students: Karjalainen Väinö, Kulomaa Vesta, Liljestrand Marcus, Peltonen Henri
Project manager: Karjalainen Väinö
Instructor: Andrei Sandru
Other advisors: -
Starting date: 12.01.2021
Completion date: 03.06.2021
Climate monitoring systems measuring temperature, air quality and other things impacting the indoor environment is something more and more people are paying attention to. Another technology that has been gaining more and more popularity is autonomous vacuums. Why not combine them? The vacuum is all the time going around the building and is therefore an ideal platform for a climate monitoring system so that we can get very detailed data from all around the building.
The high level objective of the project was to integrate an autonomous vacuum cleaner and an indoor climate monitoring system. This can be divided into smaller parts. The climate monitoring sensors should be integrated to the existing robot hardware. To enable fluent autonomous movement, a robust global path planner should be developed and the existing local path planner should be optimized. As the robot moves around and collects measurements, the collected data should be connected with location information. Furthermore, the collected data should be visualized so that a user can interpret it clearly. A remote, cloud based user interface should be created to be the visualization tool. Finally, all of these features should be connected and a working prototype built.
The robot is running on ROS and the whole system includes 3 Raspberry Pis, an Arduino, sensor box and Neato Vacuum. We combined these systems by using ROS. The robot has one of the Raspberries that acts as a master for the whole system. The box containing the sensors has one of the Raspberries and an Arduino as well as a power bank. The last Raspberry handles making the maps.
We managed to combine the two systems and the user can create informative maps based on the data the robot has collected by a single click of an icon. This requires that the user has made the map and collected data on that map beforehand but is otherwise quite easy to use. The robot’s path planning wasn’t much improved despite trying but a feature where the robot follows pre-assigned waypoints was added.