Spectrum Traffic uses Miovision Technology articial intelligence (A.I.) video analytics. Founded in 2004, Miovision has developed the most mature, advanced and most accurate A.I. video analytics available in the market today. Data collectors and governments around the world depend on Miovision's state of the art video processing to ensure that street level traffic data is counted and classified accurately.
Spectrum Traffic takes it a step further by developing its own post processing quality assurance / quality control (QA/QC) process. Once the video processing step is completed by Miovision, Spectrum extracts the data via a proprietary API, allowing our team to perform corridor and data completeness checks. While we are strong believers in A.I. technology, experience has shown us that accuracy of A.I. derived data is only as good as the humans who configure the study. Our in-house QA/QC process ensures that the A.I. achieved our minimum 95% accuracy guarantee.
Even the most advanced A.I. video analytics can't count what it can't see. That is why proper camera deployment is perhaps even more important than choosing the best A.I. when it comes to accuracy. Getting accurate data doesn't happen by accident, and it certainly doesn't happen by focusing on price alone. With over 17,000 studies completed to date, and over 85,000 hrs of video processing, Spectrum Traffic is the leader in consistently delivering accurate traffic data to its customers.
Every count conducted by Spectrum is planned by our expert staff. Our field team first reviews the site environs using Google Streetview. A "camera plan" is developed by considering lane configuration, cross-sectional width and intersection geometry. Once a camera plan is created, our team enters it into our proprietary operations platform, where the detailed spec including site specific notes is pushed to an iPad. Once in the field, our technical team are able to follow step-by-step the detailed camera plan, ensuring full camera coverage of the intersection.