May 2, 2023

Real-time mapping has become increasingly crucial, particularly as an enabler of advanced mobility services, autonomous vehicles, and smart city initiatives. However, traditional map-making methods can’t keep up with the demand for real-time data due to their high costs and labor-intensive processes. Typically, mapping involves deploying fleets of dedicated vehicles and employing manual data annotation. As a result, conventional mapping techniques are prohibitively expensive and produce outdated map layers, limiting their utility in modern, data-driven applications.

Nexar dramatically reduces the costs of real-time mapping by leveraging a crowdsourced network of camera-equipped vehicles with more than 700,000 “eyes on the road”. This network captures 94% of US roads each month and 3 billion miles of road vision data per year. Additionally, Nexar combines five key innovations to achieve this drastic cost reduction:

Crowdsourced Data Collection: CityStream Live capitalizes on crowdsourcing data from Nexar’s network of vehicle cameras. This approach eliminates the need for dedicated mapping fleets, resulting in more accurate and up-to-date map layers at a fraction of the cost.
Edge AI Processing: Nexar’s SDK, installed in-vehicle cameras and edge computing devices, enables CityStream Live to run AI models locally. This distributed AI lowers costs by reducing cloud data transfers, boosts trust through local decision-making, and crowdsources storage and processing for real-time mapping.
Change detection: Nexar’s SDK identifies differences in roads locally on the edge device, reducing the need to send all detections to the cloud. This focus on detected changes dramatically cuts data transfer, storage, and processing costs.
Frame-on-Demand: This feature allows apps and services to fetch image frames as needed using timestamps and geolocations, helping to control real-time mapping costs while providing image or video evidence.
High-Precision 3D Map Reconstruction: Traditionally, creating 3D maps required expensive equipment and specialized software. AI algorithms can now create 3D map layers from 2D images, dramatically reducing the cost and complexity. This makes 3D mapping more accessible to a wider range of organizations and individuals.
With these innovations, Nexar’s CityStream Live platform significantly reduces the need for costly data gathering, complex hardware, and manual annotation, while providing the high levels of coverage, freshness, and precision required by modern mobility services and applications.

“Nexar is revolutionizing the mapping industry by harnessing the power of AI and our vast network of ‘eyes on the road’ to create faster, more accurate, and cost-effective map layers,” said João Barros, Chief Platform Officer at Nexar. “The CityStream Live platform enables over 20 million daily detections in a cost-effective way, providing petabytes of visual data to improve safety, comfort, and mobility. Our easy-to-use APIs and SDKs empower industry players to build the next generation of mobility solutions with the unprecedented advantage of real-time mapping. As AI technology continues to advance, we’re excited to be at the forefront of this transformation, deploying AI models over the air and enabling scalable detections of relevant road objects and events at the network edge. This full-stack, end-to-end, “pixel to API” platform is already shaping a more accessible, efficient and connected future for mapping companies, automotive OEMs and beyond.”