The robotics sector is booming, yet a fundamental challenge persists: moving autonomous machines safely and efficiently from the controlled environment of a warehouse into the complex, unpredictable flow of daily life. Vancouver-based startup Ma Robot AI is tackling this issue head-on with its specialized "Embodied AI" software, designed to grant robots and autonomous vehicles human-level intuition in dense, high-traffic areas.
Founded by a unique partnership, Ma Robot AI is led by CEO Winnie Liang, who brings a background in business and workflow automation from her time at PriceWaterhouseCoopers (PwC), and CTO Dr. Mo Chen, a distinguished authority in the field as a Canadian Institute for Advanced Research (CIFAR) AI Chair and Computer Science faculty member at Simon Fraser University, with a recent visiting professorship at Stanford. They are co-founder husband and wife. The need for their combined expertise emerged during the pandemic, as they witnessed how overwhelmed hospital staff spent crucial time on manual, repeatable delivery tasks.
The Technology: Intuition Without Data
Ma Robot AI's core offering is a patent-pending hybrid AI system that fundamentally shifts how robots perceive and interact with uncontrolled environments. Unlike conventional solutions that rely on extensive data collection for specific environments (a key bottleneck for scaling), Ma Robot AI's software operates in real-time, on the edge, and requires no prior data from the target location.
The critical differentiator is the concept of "interpretable AI," which allows robots to predict how human agents will move in time and space, enabling them to navigate safely and confidently among people. This technology is applicable not just to delivery robots, but also to autonomous and computer-assisted driving systems, offering potential improvements in safety and efficiency on urban streets.
Scaling Strategy: From Integrator to Licensor
Having incorporated nearly two years ago, Ma Robot AI, currently a team of eight, is pursuing a two-phase business model. Initially, the company acts as a system integrator (SI), pairing its software with cost-efficient, high-performance third-party hardware (often sourced from Asia). This approach is currently being piloted in a BC hospital, where a robot is helping deliver lab samples-a project supported by a grant from the BC government.
The long-term vision is to transition entirely into a software licensing model. By showcasing the power of their AI through successful pilots, the company aims to establish trust with major global hardware vendors. This would allow manufacturers to license Ma Robot AI's software to upgrade their own mobile robots and vehicles, significantly accelerating market entry into high-impact sectors like healthcare and urban logistics.
Liang emphasized that Ma Robot AI will focus on AI algorithms and software, which is their core strengths, and refrain from producing robots themselves.
Eye on the Asia-Pacific Market
MA Robot AI is leveraging a Canadian Technology Accelerator (CTA) program to focus its international expansion on Taiwan's technology and manufacturing ecosystem. CEO Winnie Liang emphasized that Taiwan's strength in hardware, robotics, and supply chain efficiency makes it a perfect partner to scale their software-centric product.
The company is actively seeking strategic partnerships with local industry leaders, technology integrators, and investors, with specific interest in meeting groups like URS Robot, Turin Drive, and Kingswaytek Technology during its November visit.
Ma Robot AI is targeting an early-next-year Seed Funding Round, aiming to capitalize on the traction generated by their pilot successes and solidify their transition from a startup with promising technology to a pivotal software licensor in the global robotics landscape. Their ultimate goal is to see their AI bring robots out of the industrial warehouse and into broad, collaborative use alongside the human workforce within the next five years.
Winnie Liang, CEO of Ma Robot AI. Credit: Ma Robot AI
Article edited by Joseph Tsai
