We are building a cloud-based global infrastructure for hyperlocalization and 3D mapping via software-only enhancements to Global Navigation Satellite Systems (GNSS) such as GPS.

Technology: Urban localization using GNSS is notoriously imprecise (with errors of up to 50 meters) due to blockage of satellite signals by tall structures. However, blockage can actually be exploited by using information already available in the GNSS receiver, regarding the SNR corresponding to each satellite it sees: high SNR means that the path from the satellite to the receiver is probably not blocked, while low SNR means that the path is probably blocked (i.e., that the GNSS receiver is in the shadow of some structure). This implies that valuable (but very noisy) location information can be derived by “shadow matching” against 3D maps of the environment. At ShadowMaps, we have developed a real-time, low cost, cloud-based Bayesian localization and tracking solution that combines shadow information a...
We are building a cloud-based global infrastructure for hyperlocalization and 3D mapping via software-only enhancements to Global Navigation Satellite Systems (GNSS) such as GPS.

Technology: Urban localization using GNSS is notoriously imprecise (with errors of up to 50 meters) due to blockage of satellite signals by tall structures. However, blockage can actually be exploited by using information already available in the GNSS receiver, regarding the SNR corresponding to each satellite it sees: high SNR means that the path from the satellite to the receiver is probably not blocked, while low SNR means that the path is probably blocked (i.e., that the GNSS receiver is in the shadow of some structure). This implies that valuable (but very noisy) location information can be derived by “shadow matching” against 3D maps of the environment. At ShadowMaps, we have developed a real-time, low cost, cloud-based Bayesian localization and tracking solution that combines shadow information and raw GNSS location estimates with probabilistic motion models, to provide up to 10X improvements in urban localization. While the localization solution relies on the availability of 3D maps, the underlying Bayesian framework extends to Simultaneous Localization and Mapping (SLAM), providing a means to create 3D maps using crowdsourced GNSS data. Our technology does not require any changes in GNSS receiver hardware or firmware. It is deployed on existing mobile devices via an SDK (or via our app, or using an API call) which sends GNSS data to the cloud at regular intervals, and receive improved position information back. The GNSS data includes the estimated latitude and longitude coordinates, along with the azimuth, elevation, and SNR of each satellite in view. The ShadowMaps Bayesian framework extends naturally to incorporate additional information such as inertial navigation, WiFi and cellular.

Go-to-market strategy: While hyperlocalization using ShadowMaps is valuable when integrated into platforms such as Google or Apple Maps, we are currently targeting verticals where improved localization is an immediate need, such as car services (e.g., Uber, Lyft), delivery services, and navigation and path planning for automated and semi-automated vehicles. SDK penetration into mobile devices for localization in metro areas for which 3D maps are available, in turn, enables crowdsourced creation of 3D maps worldwide, creating a virtuous cycle for value creation.
More information

Advisors

Sanjay Srivastava
Admin
Sanjay Srivastava Advisor for strategy and bizdev

Recommendations

Sanjay Srivastava Advisor for strategy and bizdev at ShadowMaps

ShadowMaps is bringing proven research to address a large problem of urban localization. Its cloud-based solution does not need any additional hardware and is very easy to deploy to bring immediate benefits to all the applications like ride-sharing services that could benefit from significant improved localization.