September 2, 2021 | Miniature robot equipped with smart chemical sensor sniffs out buried survivors during urban search and rescue operations

02 septembre 2021 | Aider à la recherche de victimes ensevelies en milieu urbain grâce à un système intelligent de détection chimique embarqué sur robot miniature
The odor-detection module
developed and integrated by CEA-List
CEA-List, a CEA Tech institute, has developed an olfactory-signature detector and integrated it into a ground-based robot to aid urban search and rescue workers in natural and other disaster scenarios. The research was part of the EU H2020 CURSOR (Coordinated Use of miniaturized Robotic equipment and advanced Sensors for search and rescue OpeRations) project.

The main objective of the research was to develop a system capable of detecting the presence of human victims buried under difficult-to-access rubble and determining their condition to more effectively direct search and rescue efforts. The solution—an artificial nose made from biosensors with odor-binding proteins—can detect survivors’ unique olfactory signatures. The researchers had to overcome several major challenges to develop the system:

  • The device had to be miniaturized for integration into a robot small enough to be brought to the site by drone and navigate inside rubble.
  • It had to be made sensitive and selective enough not only to locate victims, but also to differentiate between living and deceased victims so that rescue efforts can be directed appropriately.
  • It had to be robust enough for use on rough terrain.
  • It had to offer long enough battery life to last for the duration of search and rescue operations.


CEA-List integrated and validation-tested the device’s vapor-phase chemical sensor module, which was built on commercially available gas sensors and innovative new sensors that use odor-binding proteins biomolecularly-engineered specifically for the chemicals of interest. These new sensors, designed to detect low concentrations of volatile organic compounds, were implemented with an artificial neural network supervised learning algorithm. The resulting real-time solution can “smell” a human victim beneath rubble and, according to the olfactory signature, determine whether the victim is still alive or, unfortunately, deceased.

The detection module integrates the sensors plus data readout, pre-processing, and wireless transmission circuits. Validation testing was completed in the lab, and then the system was piloted at a search and rescue training facility in Chambéry, France.

The focus of the research will now shift to determining how robust the device is in actual operations. Extensive field testing will be done from September 2021 to February 2022 to confirm the device’s performance and give search and rescue workers a valuable new tool.