Advanced machine vision analyzes data from multiple sensors, including lidars, cameras, and radars. Sensor fusion and artificial intelligence generate a lean and accurate representation of the objects and actors in the scene.
Core Autonomy Algorithms
Cyngn’s planning and control system evaluates the scene and all actors and then decides and executes driving maneuvers in order to achieve the vehicle’s mission
A four-dimensional, high-abstraction representation of the environment is created to predict the movement and trajectories of vehicles and people, and enable modeling of scene dynamics across use cases.
Cyngn’s on-board watchdog constantly monitors the performance of all sub-systems and reports performance indicators and events. Multiple layers of sensors and algorithms are applied to ensure robust coverage and redundancy across the system. Tele-assist systems utilize the most advanced network technologies to augment the fleet with human supervisors for any contingency.
Data Center and Pipeline
Cyngn’s system intelligently triggers event-driven data collection, which is used for developing machine learning models, and performing rigorous simulation and testing.
This data includes driving conditions data for regression testing and debugging, raw sensor data for machine vision model and algorithm development, and much more.
Cyngn’s intrinsic simulation framework, indistinguishable to the autonomous driving software from the real world, accelerates the development, testing, and validation of software upgrades and new feature deployments, through performance and event data collected from production and testing fleets.
Using accelerated run, headless operation, automatic grading, Cyngn can test orders of magnitude faster than can be achieved by road testing alone.
Cyngn’s data-informed-and-driven development process involves measuring everything. From software impact modules on hardware resources, through the performance of specific machine learning models to overall system driving capabilities and mission completion.
Insights are derived from this data to continuously improve the system and the service it provides.