The complexity of testing the performance of traffic detectors is often underestimated. The performance of such detectors is multi-dimensional and is often affected by the specifics of their installation and location, or by weather and traffic conditions. To get meaningful answers from such a test requires a well-planned logging system, highly reliable "ground-truthing" reference instrumentation and lengthy exposure to most, if not all, variants of outside conditions.

Similarly, some traffic detectors are affected by the angle of the sun, shadows created by vehicles, and other optical effects. If the test-site is on an east-west road, vehicle shadows will not affect the accuracy of counting as they would in a north-south road. On the other hand, the east-west road is more likely to reveal problems caused by the blinding of the cameras from sun-reflections off vehicles and occupancy distortions due to shadows.

While the Volume measurement is the easiest parameter to test and compare, it may not be the best indicator of performance. Most incident detection systems rely primarily on the Occupancy parameter, which is much more difficult to measure and test. There are many detector types that can count well but cannot measure occupancy.

Thus, tests that are poorly conceived or hastily conducted may lead to the wrong conclusions. Projects may fail because of a wrong selection of the most critical component - the "eyes of the system", the traffic detector.

The RTMS has been thoroughly tested by several independent organizations, the results and conclusions are summarized on the following pages: