Sanctuary AI Hires Former MDA Executive to Advance Humanoid Robotics

Sanctuary AI is pivoting from the laboratory to the factory floor. By appointing board chair Daniel Friedmann as CEO following the departure of interim leader James Wells, the Vancouver-based robotics firm is signaling an end to its primary research and development phase in favor of commercial scaling. The transition marks a strategic shift: instead of spending resources on proving that humanoid robots can perform complex tasks, Sanctuary will now focus on mass production and industrial deployment.

This leadership change arrives after a tumultuous period for the startup, which saw co-founders Geordie Rose and Suzanne Gildert depart in late 2024. The subsequent layoffs and executive turnover suggest a narrowing of focus. For investors and potential industrial partners, this move provides a clearer roadmap. It suggests that Sanctuary is moving toward its goal of deploying autonomous bots capable of independent item shuttling and high-dexterity manipulation using its proprietary robot hands.

Friedmann brings heavy-duty engineering pedigree to the role. Having led MDA Space during the development of Canadarm2, he has experience managing complex, high-stakes hardware projects where precision is non-negotiable. His background at Carbon Engineering, which successfully navigated an acquisition by Occidental Petroleum, also provides him with a blueprint for taking deep-tech innovations toward commercial exits or large-scale industrial integration.

While the robotics field remains crowded with Silicon Valley competitors, Sanctuary’s focus on 'industrial-grade' capabilities targets a specific niche: heavy labor and logistics. By moving to a scaling strategy, the company is positioning itself to compete not just on the novelty of humanoid movement, but on the reliability and throughput required by manufacturing and warehousing sectors.

What to watch next: Keep an eye on Sanctuary’s upcoming pilot programs with industrial partners to see how their robot hands perform in high-frequency, real-world environments compared to general-purpose laboratory demos.