Increasing safety and reducing cost of cloud usage for communicating weather conditions

Changes to the processing of cloud information from researchers show promise.
Changes to the processing of cloud information from researchers show promise.

Rensselaer Polytechnic Institute professor Stacy Patterson and Montana State University researcher Mike Wittie believe there is a better way for cars to process information based on sensors used to detect dangerous weather conditions.

Supported by a two-year, $250,000 grant from the National Science Foundation, these two are deep into the project entitled “A Hybrid Vehicle-Cloud Solution for Robust, Cost-Efficient Road Monitoring.”

Each year some 1.3 million vehicles are involved in accidents related to weather conditions. With cars now equipped with sensors that trigger a warning system regarding inclement weather, this real-time data seems to be a great solution. However, the cost of communicating with the cloud and having the data analyzed is prohibitive. Patterson and Wittie say creating a system in which cars can learn from each other would be more effective.

The distributive approach can be difficult because of communication glitches and breakdowns, but it is also more scalable and faster than the centralized approach, using only the cloud. It’s also more cost-effective. 

The pair is preparing to equip a small fleet of cars to prove they can accurately estimate conditions using only a portion of the data collected. This will allow them to eventually build a simulator of the system, which will warrant discussions about changing the current process.