CalmCar Vehicle Vision System

DEVELOPMENT HISTORY

2016-2019

·
Founding of company

·
Won orders from CATARC and SAIC and participated in R&D for SAIC's Project Marvel

·
Participated in NVIDIA's Inception Program

·
Sold 1st generation mass produced PDAQ to customers including SAIC, SenseTime

·
Commenced development of APA for BAIC BluePark

·
Commenced development of 1st generation of DMS, targeted at ride-hailing market

·
Commenced development of 1st gen of AVM, supporting Neta U series production

·
Obtained strategic investment from BAIC and China Unicom

·
Partnered with SlPG to develop5G-enabled L4 heavy-duty trucks for Yangshan Port Project

·
Partnered with ZF to develop driving and multi-camera parking solutions

·
Partnered with Dongfeng for high level AD projects

·
Appointed by Geely as partner for AVP solutions

·
First collaboration with overseas customer via ENTRON (Türkiye), for the development of AD functions

·
Received strategic investment from SAlC and China TransInfo
2020-2021

·
SAIC's L4 intelligent heavy-duty truck project in Yangshan Port commenced operations

·
Nominated by SAIC as series production supplier for parking solutions

·
Partnered with ZF to develop a cross-level parking solution based on visual crowdsourced 3D mapping

·
Partnered with Hozon for APA and AVP

·
Commenced collaboration with Dongfeng on visual SLAM techniques for PoC project

·
Commenced R&D of Bird's Eye View (BEV) perception based on surround-view system

·
Closed Series B fundraising

·
Obtained strategic investment from ZF

·
Became nominated supplier to VinFast for its L2+ series production models

·
Partnered with ZF and Mobileye to supply Zeekr's series production models

·
Supported SAIC Rising Auto's perception system development (based on BEV + Transformer architecture and 4D mmWaveRadar fusion)
2022-2023

·
Partnered with Mercedes-Benz to develop parking solutions for mechanical parking

·
Nominated as intelligent cabin supplier for BAlC

·
Nominated as night vision supplier for BYD

·
Nominated as Learning Automatic Parking Assist (LAPA) supplier for Dongfeng Motor

·
Developed deep learning models for Dongfeng's DMS, OMS and BSD

·
Self-developed L3 NOA solution based on vision and mmWave radar perception

·
Vehicle model installed with CC's DMS passed GSR requirements for DDAW (2021/1341). CC became the first Chinese algorithm supplier to achieve such compliance

·
Achieved mass deliveries of APA and LAPA systems for Neta S and GT

·
Mass deliveries of L2+ Highway Pilot, APA, DMS systems for VF8 and VF9

·
Partnered with ZF on APA for Zeekr, Polestar

·
Became the designated supplier for the L4 Roboshuttle project in Suzhou Taihu New Distract.

·
Commenced development of front-view perception algorithms with Dongfeng

·
Supported VinFast in obtaining qualification for RAEB NCAP
2024-2025

·
Enhanced partnership with ZF to tap into global markets

·
Mass deployed solutions for Zeekr and Polestar

·
Partnered with ZF to deliver L2.9 NOA system for top-tier domestic OEM

·
Partnered with Lenovo Connect and Zhiti Technology to deliver L4 robotaxi in multiple cities in China based on Maxus MIFA 7

·
Providing L4 service through joint ventures with local partners in several cities

·
Partnered with CRRC on Robobus global deployment

·
3 projects on night vision and CV DMS project for BYD in mass production

·
Cooperation with taxi companies for real-world driving data collection

·
Robotaxi operating license application for L4 solution commercial deployment under preparation

·
Helped CRRC electric secure a multi-million RMB Robobus project

·
Named among the 2024 Hurun China Al Top 50

·
The L4 Robobus will be launched in Wuzhong Taihu New District, expanding into the national smart mobility market

·
We established a deep strategic partnership with Horizon Robotics, a leading smart driving chip company headquartered in China, and developed a range of autonomous driving functions based on its chip platforms

SAIC Maxus

SAIC Maxus

Market: Mainland China, Hong Kong (China), Indonesia
Vehicle Models: IFA7
Functions: L3, Robotaxi

BYD

Market: China
Vehicle Models: Fangchengbao Bao 8
Functions: NV, PCW, FCW
ZEEKR

ZEEKR

Market: China, Europe, UK
Vehicle Models: ZEEKR X, ZEEKR OO1, ZEEKR OO7, ZEEKR OO9
Functions: AVM, PDC, FAPA, RP, HPP

Polestar

Market: China, Europe, UK
Vehicle Models: Polestar 4
Functions: AVM, PDC
VinFast

VinFast

Market: USA, Canada, Europe, Vietnam
Vehicle Models: VF8, VF9
Functions: AVM, PDC, DMS, FAPA, RP, R-AEB

Dongfeng

Market: China
Vehicle Models: eπ 007, eπ 008
Functions: AVM, PDC, FAPA, RP, TBA, HPP
BYD

BYD

Market: Europe, USA, Southeast Asia
Vehicle Models: Seal, ATTo 3, Qin L, Song L, CrossStar
StarFunctions: DMS

Ford

Market: China, Europe, USA
Vehicle Models: Bronco
Functions: DMS, OMS
BAIC

BAIC

Market: China, Europe, USA
Vehicle Models: ARCFOX Koala, ARCFOX Alpha T5, ARCFOX Alpha S5
Functions: DMS

CHANGAN

Market: Europe, USA
Vehicle Models: Deepal S05
Functions: DMS
Xiaopeng

Xiaopeng

Market: Europe
Vehicle Models: G06
Functions: DMS

Hozon

Market: China
Vehicle Models: NETA S, NETA GT
Functions: AVM, PDC, FAPA, RP, TBA, HPP

Panoramic Surround lmaging Rendering Technology

1. Real-time Distortion Correction for Multi-type, Multi-lens Cameras
2. Seamless Rendering of 360° Panoramic Surround lmaging, with high-simulation 3D Model Dynamic Visualization
End-to-end Perception Technology

We have independently developed a perception architecture based on Bird's-Eye View (BEV) and Transformer, which supports scalable sensor inputs. With a modular design, it can be fexibly adapted to chips with low, medium, and high computing power
Occupancy Network Technology

Semantic perception provides three-dimensional grid occupancy (3D) data of the vehicle's surrounding environment, enhancing the object recognition ability
Ultrasonic, Lidar and 4D Millimeter-wave Perception

The full-domain sensing capability based on a variety of ultrasonic sensors, lidar sensors and 4D millimeter-wave sensors
Scene Recognition and Understanding

The self-developed general road scene segmentation and recognition technology can support the recognition and output of traffic signs, road semantics, traffic rules, the boundaries of dynamic and static obstacles, etc. in both high-speed and low-speed autonomous driving scenarios
Facial Recognition

1.Based on the facial contour features, it can achieve the positioning, state perception and tracking of important facial organs in the presence of occlusions
2. The technical paper has been published in CVPR and applied to mass production projects in Europe
Visible Light and Infrared Light lmage Fusion

1. Visible light and infrared imagery data fusion to achieve higher quality night scene rendering
2. Already in mass deployment stage
Visual and LiDAR SLAM

Vision Semantic and Visual Feature Fusion-based LiDAR SLAM enable localization under a wide range of conditions

SOLUTIONS

L2-L3

CCpilot Highway

The CCPilot highway driving system is an autonomous driving solution based on the self-developed 360° surround-view perception, which supports a variety of L2-L3 level autonomous driving functions, including:

1. Adaptive Cruise Control (ACC)
2. Lane Keeping Assist (LKA)
3. Automatic Emergency Braking (AEB)
4. Lane Centering Control (LCC)
5. Navigate on Autopilot (NOA) and so on

L2-L3

CCpilot Parking Assistance

An automatic parking solution that can achieve functions such as automatic steering, obstacle monitoring, speed control, and automatic braking when the hands and feet are disengaged from the control.

1. CalmCar around-view perception system
2. Data fusion from ultrasonic sensors
L4

Next Generation Mobility

Full-stack L4 Autonomous Driving System based on Surround-view Perception and Lidar Perception Fusion.

1. Turnkey capabilities as well as fleet operation services
2. Commercial deployment in Robotaxi, Robotrucks, Roboshultle, Robobus and Robosweeper

CCCockpit

intelligent cockpit

We provide an overall intelligent cockpit solution that covering the Driver Monitoring System (DMS) and the Occupant Monitoring System (OMS), integrating multi-modal perception and interaction technologies to enhance driving safety and intelligent driving experience. The system has met international safety standards such as Driver Drowsiness and Attention Warning (DDAW) and Attention Distraction Detection Warning (ADDW), and is compatible with the Euro NCAP safety rating. It is suitable for integration into the cockpit domain controller platforms of various types of vehicles.

1. Driver Monitoring System (DMS): Monitors fatigue, distraction, and abnormal behaviors; utilizes facial recognition and eye tracking to provide safety warnings
2. Occupant Monitoring System (OMS): Monitors occupants, detects children left behind, and tracks postures; collaborates with in-vehicle systems to optimize safety and the cockpit experience
3. CCCockpit Intelligent Cockpit System: Integrates the Human-Machine Interface (HMI), voice assistant, etc.; enhances the intelligent, safe, and comfortable experience

HARDWARE/INTEGRATED SOLUTIONS

DCU/ECU

CalmCar has successfully achieved cross-platform chip adaptation (compatible with mainstream in-vehicle chips such as NVIDIA, Xilinx, Texas Instruments, and Huawei HiSilicon) through our self-developed deep learning models and machine vision algorithms, and has built a multi-modal perception fusion architecture. This solution deeply integrates data from vision, millimeter-wave radar, LiDAR, and ultrasonic radar, combined with a dual-engine decision system - which includes a rule-based algorithm framework based on high-definition maps and an innovative lightweight Al planning model. This allows for intelligent switching and collaborative operation of algorithm strategies in complex scenarios, providing full-stack technical support for autonomous driving systems

DMS

We developed our DMS (Driver Monitoring System) based on our technical advantages in visual perception and algorithms. The system uses infrared camera to detect the driver's facial features and head posture in real-time, extracting information on micro-expression changes and gaze estimation. Based on the perception results, our system classifies abnormal states and determines alert rules, issuing voice or visual prompts on the HMI (Human-Machine Interface) screen. Through an intelligent warning mechanism, our system automatically increases the sensitivity of alerts in high-speed, nighttime, and other scenarios, ensuring safe driving at all times.
Automotive-grade Camera Module

CalmCar's automotive-grade camera uses a back-illuminated image sensor, offering industry-leading sensitivity. It incorporates HDR technology, providing a high dynamic range (HDR) of up to 120dB in a highly compact automotive-grade package. This enables outstanding solutions for various Advanced Driver Assistance Systems (ADAS), including 360-degree surround perception, blind spot detection, and lane departure warning, among other features.
CalmAssistant

CalmAssistant can accurately detect vehicles ahead, lane lines, and pedestrians, providing timely warnings and alerts to reduce speed, thus minimizing collision risks. Paired with the warning display, it gives drivers a cleaner view of road conditions.