With the arrival of autonomous cars, cities will have to implement a new infrastructure. And that comes at a big cost. Cities have enough trouble managing the current one. Potholes anyone?
- Creating new infrastructure to accommodate autonomous cars can come at a high cost to cities.
- Car Pose Net offers better car tracking, using existing infrastructure and applying camera-based deep learning.
- With this cost effective solution, cities could gear up faster and more economically.
To make the transition a smoother and more economical affair, Zensors just came out with an interesting value proposition. A spinoff from Carnegie Mellon, Zensors focuses on visual sensing technologies. With its latest deep learning technology Car Pose Net, it hopes cities will choose to implement this solution.
Car Pose Net improves tracking results by fitting virtual 3D wireframes to cars, helpful especially in inclement weather and partial visual obstructions. It works a little more like the human brain, recognizing objects as shapes rather than patterns.
Here’s where city accountants’ ears will prick up. Car Pose Net works with existing city and autonomous vehicle camera systems, providing real time traffic data that’s more accurate. Car Pose Net taps into Zensor’s platform, passing data through its deep learning model and turning it into statistics that can be viewed in charts or real-time dashboards in the Zensors cloud.
Car Pose Net offers another benefit that will be attractive to city planners. Because it integrates into existing systems, it can be deployed in days or weeks. New York City will be applying the system as early as 2021 for traffic management and “congestion pricing”.
Anuraag Jain, Head of Product at Zensors also touts other applications for Car Pose Net to include traffic and parking violation enforcement. However, that’s not so exciting from a driver’s perspective.