Rock falls on remote and cliff facing roads are dangerous, hard to predict and can cause damage to vital services. In order to keep their employees aware and safe from recent rock falls the client wanted a Rock Fall Identification and Prediction System.
Powered by computer vision, the solution used dash cam footage from service vehicles to identify recent rock fall locations. Blackbook utilised our industry-leading and cutting edge machine vision expertise to build an intelligent workflow to find rock falls.
Geotagged images or videos from the client are upload to trigger the inference processing workflow. From these images and videos, GPS coordinates are extracted as well as information including date, time, vehicle identification, number of rocks, and rock size. Collated results are published to central data store, dashboard and Google Maps which is accessible by all employees.
The solution created allows the client to get accurate locations and information on the severity of recent rock falls in area regularly service by their trucks. By providing employees with this information the client is able to keep employees aware, alert and have the ability to act quickly to ensure roads are safe.