Simon is currently working closely with hardware engineers to develop a low-cost infrared sensor capable of measuring the temperature of a road’s surface. As these sensors are contactless and low in cost a network of hundreds of sensors could be quickly and affordably deployed on street furniture across a city.
Using local Wi-Fi networks along with modern web platforms and frameworks the sensed data can be transmitted, collected, and visualised in near real-time. With such a dense network of road surface temperature observations local authorities can make informed decisions about winter road maintenance. For example, the data could be used to verify the accuracy of route-based forecast models, which are commonly used to plan the salting of road networks during icy conditions.
Citizen Weather Observations
During his PhD Simon used Bayesian techniques to quantify bias and uncertainty in citizen weather observations. There are thousands of low-cost citizen weather stations submitting near real-time data to data hubs such as the Met Office’s WOW website. Simon developed a framework through which citizen air temperature observations could be passed. The framework gradually learnt observational bias and uncertainty over time. Applications wishing to use such citizen observations would then have a metric on which to base their confidence in the data.