Company attributes
Other attributes
Phantom AI is developing advanced driver-assistance systems (ADAS) for facilitating autonomous driving vehicles and driver safety. The company is developing software stacks and sensor suites for autonomous driving while providing various levels of ADAS and autonomy in those solutions. This includes using an optical sensor suite and computer vision technology with deep-learning-based detection and geometry-based filtering systems. Phantom AI's solutions are developed to meet current original equipment manufacturer (OEM) autonomous requirements and to push towards greater levels of autonomy.
Phantom AI was founded in 2016 by CEO Hyunggi Cho and CTO Chan Kyu Lee while the two worked at Tesla and shared an interest in the possibilities of ADAS technology and its capability to bring autonomous technology to the market before fully driverless cars will reach market. Phantom AI's long-term vision is to develop autonomous technology through ADAS technology by introducing creeping levels of automation, starting at L2 or L3, a few levels below true self-driving, while developing towards L4 or L5 full automation systems.
Phantom AI's PhantomVision is a computer vision solution using deep learning and is developed to be scalable and flexible, compliant with Euro NCAP ADAS features. It is built to support vehicle, pedestrian, bicyclist, free-space, traffic sign, and traffic light detection. The solution is developed for visual perception, to enable single or multiple cameras to autonomously recognize objects on a road and traffic directions in its vicinity and driving path. The system can be used with a combination of front, side, and rear cameras to eliminate blind spots and increase the system's safety margins. This can be used to trigger ADAS functions such as auto emergency braking, adaptive cruise control, traffic jam assist, and lane-keeping assist systems.
Phantom AI's PhantomFusion is a platform-independent sensor fusion and object-tracking system. The system is developed to use a modular sensor fusion solution to create an environmental model through various sensors, such as cameras, radar, LiDARs, and ultrasonics, to create accurate and clear detections and to guarantee detection in the case of partial sensor fails by combining strengths of various sensors.
