How Road Weather Forecasts Can Enhance Consumer Navigation Systems


In this modern era of convenience, once standard methods of navigation have been upgraded by GPS systems.This technology can make our lives easier and safer, but could the integration of GPS systems with hyper-local road weather forecasting bring convenience and safety to another level? GPS systems are certainly useful for finding locations and avoiding traffic congestion, but are not so effective at avoiding dangerous road conditions that put motorists’ lives at risk. Extreme weather contributes to a significant volume of road accidents and poses severe safety concerns for drivers. Inclement weather leads to at least 1.2 million vehicle crashes each year, accounting for around 21% of all accidents.

GPS navigation systems have primarily drawn weather data from atmospheric forecasting models. While atmospheric weather forecasts give broad ideas of the weather of a region, conditions on the road often differ from the weather in the atmosphere above. For example, roads can melt falling snow or hold onto icy conditions, even in sunny weather.

This often leads to drivers grasping to determine what the actual condition of the road might be. A forecast might predict rain, but if the ground is at a sub-freezing temperature this can generate black ice that motorists aren’t prepared for. As Charles McGill of the National Oceanic and Atmospheric Administration puts it, “Forecasting is still an inexact science.”

This is where road weather forecasting technology can play a role, by enhancing consumer navigation systems and providing accurate, real-time information to drivers. Fathym’s Ground Truth™ road weather forecast captures observational data from a vast network of environmental sensors and blends it with a multitude of atmospheric forecast models. The statistically-based model employs machine learning to create a true road forecast rather than extrapolating road conditions purely from the weather above.

Physical weather models are data driven. They require more raw input for training on a specific forecast point (road type, road base, heat transfer, shade, UV), which yields slower processing cycles and limits scale across cloud resources. Although the physical model will provide a more accurate forecast at one specific point, it loses its value as the user scales across thousands or millions of points along a road segment.

In contrast, Fathym’s statistical model consistently outperforms the physical model in this use case. Advanced machine learning and model training techniques produce high-resolution, scalable road weather forecasts. When integrated with consumer navigation systems, drivers can dynamically re-route based on forecasted and changing conditions, and accurately update estimated time of arrival. Driving will always involve a certain degree of risk. However, hyper-local road weather forecasting can give motorists the means to navigate with more security, regardless of what the weather throws at them.