Search-and-rescue robotics

In the first seventy-two hours after an earthquake, the chance of finding survivors under rubble drops dramatically with each passing hour. Human rescuers are extraordinary — but they cannot fit through a gap the size of a drainpipe, they cannot safely enter a building actively collapsing, and they cannot be everywhere at once. A robot can be sent in first. That idea drives search-and-rescue (SAR) robotics.

SAR robots operate in what researchers call unstructured environments — settings where nothing is where you expect it to be, the floor may give way, and the air may be full of dust, smoke, or chemical vapour. This makes SAR robotics one of the hardest problems in the field. A factory robot works in a world designed for it. A SAR robot works in a world that has just been violently unmade.

What the robots actually do

SAR robots are deployed for three main tasks. Search involves mapping the interior of a collapsed structure — using cameras, thermal sensors, and acoustic detectors — to find signs of life before humans enter. Reconnaissance means characterising the hazard: measuring air quality, detecting radiation, or assessing structural integrity. Access involves physically reaching a survivor: passing through a narrow gap, descending a flooded shaft, or navigating rubble that would shift under a human's weight.

Ground robots for SAR are typically tracked, like tiny tanks, because tracks handle rubble better than wheels. Aerial drones map disaster zones quickly from above, creating a picture that guides ground teams. Aquatic robots handle flood rescues and ship-wreck searches.

The Fukushima deployment

When the Fukushima Daiichi nuclear plant was damaged by the 2011 tsunami in Japan, radiation levels inside the reactor buildings were lethal to humans within minutes. iRobot's PackBot and Warrior robots — originally designed for military bomb disposal — were sent in to map radiation levels, inspect damage, and assist in the initial response. They operated in conditions where any human would have received a fatal dose in under an hour. The deployment demonstrated both the potential and the limitations of SAR robots: the machines provided irreplaceable data but required constant operator intervention and sometimes became stuck in debris.

The honest challenges

Robin Murphy, a leading SAR robotics researcher, has documented that in most real disaster deployments between 1995 and 2020, robots played a supporting rather than a decisive role. Communications break down inside rubble. Batteries die. Robots get stuck. Human rescuers, who can improvise and feel their way through ambiguity, are still more effective in many scenarios. The DARPA Robotics Challenge in 2015 was partly designed to push the field forward — and its results were humbling: robots took many minutes to open a door that a human does in two seconds. Progress since then has been steady, but SAR robotics is an area where the gap between laboratory performance and field performance remains large.

The most important unsolved problem in search-and-rescue robotics is not finding survivors — it is communicating reliably through three metres of concrete rubble.