A robot that can be guided by an electronic nose invented by a postgraduate student could aid search and rescue missions following natural disasters.
During her postgraduate stay at the Tecnológico de Monterrey (ITESM) in Mexico, Blanca Lorena Villarreal developed a device that allows multiple robotic platforms to follow the path of certain odours
The olfactory system is based on artificial intelligence algorithms that enable the detection of the scent of alcohol, but with some modifications to the system and the algorithms it can be made to recognise other chemicals and odours such as blood, sweat or human urine.
In the first phase of development, the student wondered about how living things carry out the process of odour recognition and then transferred that knowledge to the mathematical sciences, to create a series of algorithms.
"We note that, biologically, animals perceive the direction of an odour using two characteristics: it comes at different concentrations to the nostrils, and, because it is appreciated with a time difference. These two factors can identify from which direction a certain aroma comes," said Lorena Villarreal
This is how chemical sensors that mimic the nostrils, and which are separated by a septum, will perceive specific odours. The data, said Lorena Villarreal, is sent by radio to a computer, where it is analysed in real time to know the origin and direction of the aroma, using programmed algorithms.
"Unlike other olfactory systems, this has the feature that in each cycle of ventilation the air chamber empties, making sensors ready for a new measurement," she added. This means the technology takes only one cycle to detect that there has been a change of direction in the path of smell, which enables the robot to perform the tracking faster.
The olfactory system has been integrated with a robotic platform funded by CONACYT, Mexico’s national science council, to achieve test the device for deployment to hypothetical emergency zones.
Lorena Villarreal is further developing algorithms that allow the discrimination of odours, to give the robot some artificial intelligence that contributes to decision making processes.