


Smart City or AI City
Over 55% of the world population lives in urban areas. This trend is expected to grow putting more burden on road networks. Technology can help manage this better with real time visibility to vehicular and pedestrian behavior at intersections, corridors and the traffic network as a whole.
Using our models, delivered as API or SDK on edge; you can gather data from intersections to infer signal state, saturation flow rate, saturation headway, Purdue split failure, red and green occupancy ratio and arrivals, simple and average delay, arrival volumes and pedestrian count and delays. The models can also detect instantaneous velocity, erratic driving, red light violation, cross walk encroachment, accidents and susceptibility to accidents, spillage, road hazard, intersection behavior and model traffic flow across city traffic for city planning and zoning.
Safe driving
Autonomous cars and trucks will make roads safer in the future. But we have the technology today to make driving safer using technologies designed for Self driving cars. Our models can be deployed in cars, trucks with any dash-cam capable of streaming video to our edge (Nvidia Nano) can warn drivers of forward collision, tailgating, distracted driving and many other unsafe behaviors all using Nvidia Nano. Unlike smart dash-cams, we don’t need a hard braking event to record unsafe driving. Our models also work without any road markings opening up the potential to new geographies. The models are robust and the camera can be placed anywhere as long as it has 90% clear view. The models also use relative velocity to reduce false positive.
Smart security
Security cameras are everywhere. Lower cost, better quality, and wireless access have increased the security cameras footprint to more homes, commercial locations, streets, and public places. While installing, connecting and processing video has improved, the insights we derive from the video has not. This has resulted in security cameras used primally as a deterrent and post event analysis.
Using our models, on edge we can connect to any legacy camera source capable of streaming video. Our models are robust and offer the ability to draw polygons on areas of interest and not just rectangles. The model also offer filters to be notified against such as objects, people or behavior including temporal you are interested in. All this can be delivered on edge with Nvidia Nano.