At CES 2026 in Las Vegas, Singapore startup Strutt debuted ev1, an AI‑powered smart wheelchair that can autonomously navigate open spaces, avoid pedestrians and follow voice commands, according to a January 17 report relaying Straits Times coverage. The device uses LiDAR, 25 environmental sensors, cameras and large language models like ChatGPT to interpret multi‑language voice requests, with an expected price of about $7,499 and deliveries planned for Q2 2026.
This article aggregates reporting from 1 news source. The TL;DR is AI-generated from original reporting. Race to AGI's analysis provides editorial context on implications for AGI development.
Strutt’s ev1 illustrates how frontier-like AI capabilities are quickly embedding into physical assistive devices. This isn’t a research robot in a lab; it’s a wheelchair designed to operate in messy, human environments like hospitals and public spaces, using LiDAR, SLAM and LLM-powered voice interfaces. That combination—autonomy, perception, and natural language control—looks a lot like an early template for the ‘AI embodied agent’ many AGI roadmaps anticipate.
From a competitive standpoint, this shows how smaller startups in Singapore and the broader ASEAN region can build differentiated products by composing global foundation models with strong robotics stacks and local domain expertise. Strutt doesn’t need to train its own GPT‑scale model to deliver value; it can lean on services like ChatGPT for language understanding while focusing on safety, hardware integration, and regulatory approvals.
In the AGI context, deployments like ev1 will provide valuable data on how humans interact with semi-autonomous AI in safety‑critical contexts, especially for vulnerable populations. Those lessons—about trust, override controls, and failure modes when navigation and language systems disagree—are directly relevant to designing higher-capability agents later. The labs and companies that learn fastest from these early physical deployments will have an edge in making future AGI-powered robots socially and operationally acceptable.



