“Prevention is better than the cure,” says Dr. David Lary. As a professor of Physics at the Hanson Center for Space Science at the University of Texas at Dallas and Founding Director of MINTS (Multi-Scale Integrated Intelligent Interactive Sensing), Dr. Lary has over 30 years of experience applying science to solve real-world problems and support societal needs. One of these real-world problems is detecting methane gas. Data shows that the number of gas explosions in the Dallas-Ft Worth area has increased since 2004. Sensors can continuously monitor methane levels and detect urban gas leaks immediately. With public safety at risk, they can offer precise locations of the sources, allowing for quick and targeted responses by emergency services.

Dr. David Lary at UTD @ the Richardson IQ®.

Expensive sensors, like those used in advanced scientific measurements, offer unparalleled accuracy but come at a steep price. On the other hand, low-cost sensors are affordable and accessible but often suffer from inconsistency and reduced precision. To bridge this gap, Dr. Lary and his team, as part of the OWL project, are leveraging machine learning to create “Software Defined Sensors.”

Example Sensor.

The process involves placing these low-cost sensors alongside a high-quality reference sensor in a controlled chamber where variables like temperature, humidity, and methane concentration can be adjusted. This reference sensor provides precise readings and serves as a benchmark. The team then uses machine learning to model the relationship between the reference sensor and the low-cost sensors, calibrating them based on environmental conditions such as temperature, pressure, and humidity.

Dr. David Lary and MINTS are located within the Richardson IQ HQ. UTD @ the Richardson IQ® is an innovative space fostering corporate innovator and university collaborations, supporting Richardson and the region’s startup and entrepreneur community, and advancing UT Dallas research. Dr. Lary has fostered great collaborations with the many people who come through the facility. He looks forward to expanding his work with the OWL project by utilizing more machine learning to wrap a sensing suite with smart software. Continuing to help public safety will enhance the accuracy and capabilities of the air quality sensors for methane and carbon dioxide by calibrating the low-cost sensors against reference sensors within the calibration chamber.