Hyperedge- . IoT, Embedded Systems, Artificial Intelligence,

Firefighting is a dangerous profession, but it is possible to mitigate some of that danger with good data. When firefighters entered a burning building, their biggest fear is the unknown. They don’t know if they can trust the structural integrity of the building, if there is a pocket of toxic or explosive gas, or how to navigate the interior to find casualties. As part of a project called HelpResponder, a team of researchers from Universidad Rey Juan Carlos and Universidad Autónoma de Madrid created a robot that can enter a building to gather the data firefighters need to do their job safely.

This robot, which is a mid-sized rover, can operate via manual control or in an autonomous mode. In both cases, its job is to explore buildings, either during a fire or after a disaster, to map the interior and find hazards. Its camera system allows for visual detection, but it also has a host of integrated sensors for detecting elevated temperatures, gas pockets, and more. With that information, firefighters can then enter the building and rescue anyone trapped inside while avoiding hazardous areas or bringing the equipment necessary to deal with them.

Hyperedge- . IoT, Embedded Systems, Artificial Intelligence,

Control and monitoring happens on two levels. At the high level, a Raspberry Pi 4 Model B single-board computer records video, handles mapping operations, and coordinates autonomous navigation. At the low level, an Arduino UNO WiFi Rev.2 collects incoming sensor data and controls the motor driver. The onboard sensors include a temperature/humidity sensor, an air quality sensor, and ultrasonic sensors for navigation. Thanks to a modular design, additional hardware can be added to fit specific scenarios.

The team plans to continue improving the robot, particularly its autonomous operation mode. But they’ve already tested it in simulations and the real world with positive results. More details on the HelpResponder can be found in their paper here.

Image credit: Fernández Talavera et al.

Read more about this on: Arduino Blog