FireBot

Autonomous firefighting home robot intended to reduce response times to fires.

Real-scale fire testing at the firehall. Kudos to the firefighters who helped make this possible!

InspoScience IRIC 2020 Presentation

Slideshow | InspoScience

Abstract

Although firefighting has improved significantly with advances in telecommunications, refined training and experience, major risks posed to both firefighters and victims remain, especially as high-rise buildings, synthetic materials and lighter, more flammable building materials continue to become increasingly prominent. Focusing on on-site technologies can significantly reduce latency of response. Currently, established fire safety technologies intended to be installed and to remain on-site, mainly sprinklers and smoke detectors, are unaware of their surroundings and designed to notify victims and first responders solely of the presence of a fire, thus depending almost entirely on the arrival of first responders. Even after first responders arrive, procedures that must be carried out before extinguishing of the fire occupy precious time. A novel solution is proposed and prototyped, with emphasis on responding to a fire as soon as possible, during its incipient stage, and replacing the need for humans to be physically present to manage the fire. In case of a fire, the robot promptly recognises the fire through visual means, alerts firefighters and victims electronically and attempts to extinguish the fire autonomously. The robot proposed will preserve lives of civilians and firefighters as well as the trauma that entails the loss of a loved one, ultimately denying Death and Grief from hindering people’s lives.

Method

The main objectives for the robot are alerting first responders and victims of the structural fire, providing firefighters with the ability to assess the fire from within the area concerned and attempting to extinguish the fire, all during the incipient stage of the fire. To achieve almost immediate response time, the robot resides in the area concerned, such as a residential, commercial or office structure, and is therefore relatively small. Through optical and near infrared vision and machine learning techniques for image recognition, the robot is able to detect fires with greater precision and earlier than smoke detectors. Further, it is able to collect valuable information such as video in real time, instantly engage in two-way communication with firefighters, and provide the data collected. Next, the robot automatically begins to autonomously extinguish the fire by dispersing fire extinguishing agents such as water, driving and rotating according to the location of the fire as well as calculations of the trajectory to be dispersed. At any time, certified personnel can take over the robot via remote control, with access to live video from the robot’s cameras, its driving functionalities and rotation of the cameras and nozzle.

Stack

Tank Drive

DC Motors controlled by Raspberry Pi GPIO via an H-bridge

Aiming: 2-Axis Rotation

2 servo motors controlled by Raspberry Pi GPIO using PWM (pigpio library) (with a swivel plate for panning)