Kount: Powering Computer Vision for IoT

Just over two months ago, we initiated a pilot to analyze vehicular traffic in Buenos Aires, Argentina using Kount, Kaaenaat's solution for performing AI & Computer Vision solution for IoT devices at the edge. 

Today, I'm happy to report that our pilot has completed successfully and discussions are under way for it to be more widely deployed at traffic intersections across more cities.

In addition to counting vehicles, Kount performs additional analysis such as categorization of vehicles based on vehicle type, speed & direction of movement, identify traffic violations, accidents, traffic jams and more.

Kount can also detect anomalies in traffic flow, such as due to an accident, road repair, or an event such as a rally, concert, sports meet or a storm.

Currently, Kount runs #DeepLearning and #ComputerVision algorithms on #IoT devices powered by ARM-Cortex & Qualcomm Snapdragon chipsets (the same chips that power most Android mobile phones). Since these #IoT devices are located in close proximity to the camera capturing the data, this eliminates the need for video streams to be sent to the cloud and therefore the need for expensive fiber optic connectivity, mobile data costs & high maintenance & support costs that come along with them.

Kount's technology typically provides total cost of ownership (TCO) savings of over 87% compared to current offerings in the market.

Yeah, we think that's pretty amazing too!

We are now expanding the gamut of our solution to other use cases such as retail spaces, transportation hubs, exhibitions & trade shows, sporting & entertainment venues, parking lots and malls.

If your firm or client can benefit from real-time analytics of vehicular or people traffic, identifying patterns & anomalies in traffic, or simulating the impact of changes such as congestion pricing, store redesign, change in real estate rentals or extreme weather, do reach out to us at [email protected]. We'd love to hear from you!