Lyte 的旗舰系统 LyteVision 主打把多种感测融合到单一平台：包含相机、惯性运动感测，以及可同时量测距离与速度的 4D 感测器，将位置与影像等资料整合，产生可立即用于决策的输出。公司表示将用既有资金强化核心产品、扩编人力并扩大营运，并预期在未来 3～5 年于「安全」面向展现有意义的进展；其技术也在今年 CES 前获得机器人类别的创新奖。
在产业面，AI 机器人市场被预估到 2030 年将达 1,250 亿美元，但 McKinsey 指出有 60% 的工业公司缺乏导入机器人自动化（含感测整合）的内部能力。Lyte 的策略是自研晶片、光学与软体，提供感测与元件的即插即用方案，以缩短常见可长达数年的整合周期；公司目前约 100 名员工，应用场景涵盖人形机器人、移动机器人与 robotaxi。
Former Apple Face ID leaders have launched Lyte in Mountain View, aiming to become a “visual brain” for robots. Founded in 2021 by ex-Apple employees Alexander Shpunt, Arman Hajati, and Yuval Gerson, Lyte emerged from stealth in early January 2026 after raising about $107 million. Shpunt previously co-founded PrimeSense, acquired by Apple in 2013 for $350 million, and the firm also powered Microsoft’s Kinect.
Lyte’s flagship platform, LyteVision, combines three sensor types—camera, inertial motion sensing, and a 4D sensor that measures distance and velocity—then fuses location and visual data into one system for immediate, actionable outputs. The company says it will deploy existing capital to deepen the core product, hire, and scale operations, targeting measurable safety progress over the next three to five years; the technology also received a CES Innovation Award in robotics ahead of this week’s show.
Market signals point to fast growth and a capability gap: the AI robotics market is projected to reach $125 billion by 2030, yet 60% of industrial companies lack the internal ability to implement robotic automation, including sensor integration, per McKinsey. Lyte plans custom silicon, optics, and software to reduce vendor selection and shorten integrations that can take years, offering a plug-and-play stack. With roughly 100 employees, it positions the tech for humanoids, mobile robots, and robotaxis.
