Momenta is building the “brain” for autonomous driving. As software partner, we provide OEMs and Tier 1s with safe, reliable, and mass production-ready software of multi-level autonomy for different driving scenarios.
Our products include modular products and solution products.
As mass production data, data-driven algorithm and closed loop automation accumulate and iterate, the Flywheel will pick up speed and enter a period of explosive commercial growth.
As mass production data, data-driven algorithm and closed loop automation accumulate and iterate, the Flywheel will pick up speed and enter a period of explosive commercial growth.
MSD (Momenta Self-Driving) is the full autonomy solution designed for open roads. It is built on the basis of strong system engineering and algorithm development, such as multi-sensor perception, HD Map, and path planning.
front view side view surround view rear view
3 4 4 1
FOV30°/50°/120° FOV100° FOV180° FOV50°
lidar
1
ultrasonic sensor
12
front radar corner radar
1 4
Mpilot is an autonomous driving software that can work with low-cost and mass production ready sensors and SoCs. It offers maximized continuity of autonomous driving experience covering highway/city expressway, parking, and urban scenarios.
Mpilot Highway, integrated with HD Map, is a world-leading autonomous driving software solution created to provide the best possible driving experience, save driving time and improve safety on road.
Mpilot Highway, integrated with HD Map, is a world-leading autonomous driving software solution created to provide the best possible driving experience, save driving time and improve safety on road.
Mpilot Highway, integrated with HD Map, is a world-leading autonomous driving software solution created to provide the best possible driving experience, save driving time and improve safety on road.
Ramp-to-Ramp Driving
HD Map and localization allow us to know exactly where we are in complex driving scenarios such as on/off ramps. With accurate real-time perception of dynamic surroundings and historical trajectories of crowdsourcing vehicles, we do dynamic path planning with high confidence to automatically enter and exit ramps.
Automatic Lane-Changing
The vehicles can act automatically when they need to change lanes like an experienced driver. With the help of HD Map and localization, and 360 degrees perception, the system initiates lane changes under proper situations.
Driving Behavior Prediction
under Cut-In Scenarios
With the multi-camera fused perception system, we can better predict the cut-in intentions of surrounding vehicles and handle them timely and comfortably on highway.
front view side view side view surround view
front windshield rear-looking front-looking surroundings of car
2 2 2 4
FOV50°/120° FOV100° FOV100° FOV185°
front radar corner radar corner radar
front side rear side front
1 2 2
Mpilot Parking provides enjoyable intelligent valet parking experience via superior surround view perception, high precision in-door mapping and localization technologies.
Mpilot Parking provides enjoyable intelligent valet parking experience via superior surround view perception, high precision in-door mapping and localization technologies.
Mpilot Parking希望为用户带来每天可解放驾驶员30分钟的智能泊车体验,无需人员驾驶实现单车智能自主泊车,实现跨层泊车、智能绕行、避让车辆行人、在线寻位等多项功能。
Smartly handle complex traffic and
pedestrian situations
Dynamically detect parking-spot. Support dedicated or random parking spot
Automatic vision mapping. High accuracy.
Crowdsourcing
4 * surround view cameras
1 * front view camera
12 * ultrasonic sensor
1 * consumer grade IMU
1 * consumer grade GPS
Mpilot ICP (In-Cabin Camera Perception) is a deep learning based high-performance and automotive-grade perception software for interior vehicle environments. The Driver Monitoring System offers better driver safety for autonomous driving. The Occupancy Monitoring System extends HMI experience to more occupants in cabin.
Mpilot ICP (In-Cabin Camera Perception) is a deep learning based high-performance and automotive-grade perception software for interior vehicle environments. The Driver Monitoring System offers better driver safety for autonomous driving. The Occupancy Monitoring System extends HMI experience to more occupants in cabin.
DMS基于深度学习提供低成本、高性能、车规级的DMS解决方案,有效降低交通事故率,极大提高驾驶安全性,为驾驶员出行保驾护航。
Face recognition
Fast and accurate recognition of driver’s identification, age, facial expressions, etc.
Drowsiness detection
Monitoring of closed eyes and yawning to remind drivers to take some rest, thus to reduce accidents on the road.
Distraction detection
Head pose and gaze tracking to remind drivers to pay attention to roads and improve driving safety.
Action detection
Detection of dangerous driving actions, such as smoking, yawning, making phone calls, etc., to ensure driving safety by sending out dangerous action warnings.
High gaze accuracy < 2°
Less than 2° gaze tracking accuracy.
Over 1 billion km data
Accumulated over 1 billion km of complex scenarios for deep-learning model training.
Proprietary ground-truth system
Less than 0.05mm 3D reconstruction accuracy of our proprietary ground-truth system.
Mpilot Front Camera Perception offers superior front-looking vision perception and high-precision localization that meet the requirements of Euro NCAP and Advanced Driver Assistance Systems (ADAS).
Mpilot Front Camera Perception offers superior front-looking vision perception and high-precision localization that meet the requirements of Euro NCAP and Advanced Driver Assistance Systems (ADAS).
Mpilot Front Camera Perception offers superior front-looking vision perception and high-precision localization that meet the requirements of Euro NCAP and Advanced Driver Assistance Systems (ADAS).
Superior Performance
Extreme performance optimization across a variety of mass production SoCs
Wide-Adaptability
Trained on more than 1 billion km of data, robust algorithms under various road, weather, and lighting conditions
Localization Feature Included
10cm localization accuracy achieved via front camera perception and HD Map
HD Map for autonomous driving entails features of precision, freshness and scalability. 3D positioning of the surrounding environment is reconstructed by extracting semantic information from 2D images and fusing with GPS and IMU data. “Live” map which powers various levels of autonomous driving is automatically generated from crowdsourcing vehicles with low-cost and mass-produced sensors.
HD Map is used by different autonomous driving software modules such as localization, planning and control.
HD Map for autonomous driving entails features of precision, freshness and scalability. 3D positioning of the surrounding environment is reconstructed by extracting semantic information from 2D images and fusing with GPS and IMU data. “Live” map which powers various levels of autonomous driving is automatically generated from crowdsourcing vehicles with low-cost and mass-produced sensors.
HD Map is used by different autonomous driving software modules such as localization, planning and control.
HD Map for autonomous driving entails features of precision, freshness and scalability. 3D positioning of the surrounding environment is reconstructed by extracting semantic information from 2D images and fusing with GPS and IMU data. “Live” map which powers various levels of autonomous driving is automatically generated from crowdsourcing vehicles with low-cost and mass-produced sensors.
HD Map is used by different autonomous driving software modules such as localization, planning and control.
Crowdsourcing based HD Map enables low cost, high precision and robust localization.
Crowdsourcing based HD Map enables low cost, high precision and robust localization.
Crowdsourcing based HD Map enables low cost, high precision and robust localization.