GlobVision design and developed advanced algorithms for modeling subsurface soil temperatures measured by probes, and estimating missing values for better frost monitoring beneath roadways.
In cold climates like Canada, frost reaching down below roadways can cause significant damage to the pavement in the form of non-uniform heave and cracks during the freezing and thaw cycles. Therefore, it is critical to measure subsurface temperatures to monitor frost conditions beneath roads to regulate maximum truck loads that can pass over the road. Missing or corrupted probe readings interrupt frost monitoring.
Critical to this application was the development of innovative data processing methods for reliable detection of corrupted or erroneous temperature readings. The mathematical models for temperature evolution at each depth layer of the soil were based on data-driven nonlinear regression methods, taking into account environmental conditions like ambient temperature and dew point temperature.
The models and algorithms GlobVision developed enable uninterrupted frost monitoring and allow for more informed decision making by load regulatory bodies.