A Test of Precision: Seyond’s SIMPL-Intersection Solution vs Inductive Loops in Tampere, Finland

In the heart of Tampere, Finland, Seyond recently put its intersection management platform called SIMPL-Intersection to the test, aiming to validate the accuracy of its vehicle count data against the existing inductive loop system. This test was a critical step in demonstrating the capabilities of SIMPL-Intersection, which integrates LiDAR technology and AI-powered perception (v1.3) to offer a cutting-edge solution for traffic management.

The test was conducted across five lanes over a span of four days, between June and July 2024. The Seyond team compared the vehicle count data from SIMPL with that of the traditional inductive loop system. Surprisingly, our test uncovered a major problem with the inductive loop's accuracy while SIMPL-Intersection proved to be the most accurate solution when compared to a manual count.

The results.

After 4 full days of data, SIMPL's data closely matched that of the inductive loops, with only minimal deviations noted for four of the five lanes.

However, when the team examined the data from lane D-1, they encountered a significant discrepancy. SIMPL’s counts differed from the inductive loop data by 39% to 49% depending on the days.

Determined to uncover the cause, the team decided to take a closer look.

They downloaded 24 hours of LiDAR perception video feed from SIMPL and assigned an expert to manually count the vehicles passing through lane D-1. The results were telling: the expert recorded 137 vehicle events, while SIMPL counted 134. The inductive loop, however, recorded a significantly lower count, revealing a major problem with the loop's accuracy.