We primarily develop permissively-licensed, open software for robotic data processing. Our pioneering multi-sensor, non-Gaussian, multi-agent, and cloud enabled state-estimation software solves the robust localization, mapping, kinematic, data association, and semantic understanding problem. By leveraging the vanilla (open) software, our clients can accelerate their own proprietary software stack development by focusing on the task-specific aspects of their problem. Our solution provides a pathway to leverage value from the community, our team’s navigation expertise, and thereby reducing the cost-of-ownership and new technology risk. To further reduce cost, we offer our software-as-a-service via the cloud with pioneering support for multi-agent and multi-session applications, including heterogeneous hand-held or robotic devices.
We focus on multi-sensor applications where understanding of ambiguous data and spatial context is critical, and on systems where humans can benefit from computational autonomy. Our approach allows seamless integration between Machine Learning and statistical Inference as a combined A.I. — humans bring good judgement while computers do the work.
As and when sensible, we also help develop proprietary software for our customers as non-recurring engineering (NRE). Our upcoming subscription models provide technical support surrounding the open software, including bug fixing, feature requests, hosted compute, and enterprise services.
Our software license does have one restriction, prohibiting 3rd party use for a cloud offering, but we do offer more conventional software licensing terms for these larger cases — please contact us at email@example.com for more details.