Nexar is a leading AI vision company paving the way for a safer, more efficient driving experience.
At the heart of our mission is to improve road safety, harnessing the power of advanced technology together with a comprehensive network of connected cameras on the road. We understand the transformative impact of digitizing the world, and we're dedicated to applying technology to create social welfare.
Our network of connected cameras and our leading vision-based AI enables us to see everything that cars can see, offering the ability to crawl the physical space. Camera users, by trusting us and sharing anonymous data, contribute to the development of services that benefit not only their own driving experience but also other drivers.
For instance, our cameras are able to detect road events such as work zones. These are sensitive areas, especially if there is a lane impact that affects the typical driving behavior and traffic. Leveraging our vision network, drivers can be informed about the road dynamics that affect their route and take actions that could save time, and money and reduce risks. Such information is also useful for road management entities to better monitor and inspect the respective work zone. Other examples relate to our ability to detect road signs, potholes, and debris. The latter can be particularly dangerous and time-sensitive, so its immediate detection and communication through our network and services enables us to provide warnings to nearby drivers, making them more alert and cautious.
While building such services, privacy has always been our utmost concern. We will never share individual data with any third party without explicit authorization. We adhere strictly to legal and regulatory standards while ensuring our technology cannot be misused for individual tracking by anyone.
Our vision network is built with state-of-the-art anonymization techniques designed to secure your privacy and the privacy of bystanders. That is why we remove all 1st and 3rd party Personal Identifiable Information that might be captured by the road-facing camera before we send any data to our cloud. To accomplish that, we have trained a multi-task on-device Neural Network model to detect identifiable elements, such as faces, pedestrians, license plates, and vehicle dashboards. Once detected, we apply Gaussian blur around the area of the identified elements producing the anonymized data. This guarantees that the shared data that reaches our cloud is fully anonymized and non-identifiable. Finally, we apply additional measures to limit the scope and amount of data collected, to reduce the risk of tracing any anonymous data back to an individual.
We all see that the future will include more and more cameras and sensors in the physical world, and that is why it is imperative that we deploy such programs with utmost care to privacy, to gain the societal value of a collision-free world, without suffering any consequences to our free society.