While overshadowed by robotics giants like Boston Dynamics, the Canadian firm Sanctuary AI has been steadily advancing its line of humanoid robots dubbed “Phoenix.” This week, Sanctuary introduced the 7th generation iteration of Phoenix, highlighting the system’s human-like movements and impressive speed at learning new tasks.
Rather than focusing on the robot’s full-body locomotive abilities often showcased by competitors, Sanctuary’s new videos Zero in on the Phoenix’s dexterous torso and arm movements as it deftly sorts products by color. The company claims its latest humanoid offering can automate entirely new tasks in under 24 hours through advanced machine learning capabilities.
“With generation seven, we have a system that we believe is the most closely analogous to a person of any available,” stated Sanctuary AI co-founder and CEO Geordie Rose. “We see this as not only the cornerstone of general-purpose AI robotics but a critical step on the path to artificial general intelligence, and we’re thrilled to be leading the charge on it.”
While the demo shows relatively simple object sorting in a structured setting, Rose claims this rapid task learning represents a key milestone toward the long-sought goal of general artificial intelligence that can flexibly adapt to different scenarios akin to humans.
Sanctuary has been deploying prior Phoenix iterations in real-world manufacturing environments, including a recently announced deal to bring its robots to Magna’s auto facilities. The 7th generation model introduces further refinements like increased uptime, expanded range of motion, lighter weight, and optimized cost through an improved materials bill.
Even as eye-catching robotics videos from firms like Boston Dynamics continue grabbing headlines, Sanctuary’s under-the-radar progress in humanoid dexterity and task learning highlights an oft-overlooked frontier in the quest to make robots more versatile, intelligent assistants in industrial settings and beyond. As the field inches toward general AI, companies advancing capabilities around robots seamlessly learning complex new skills could prove pivotal.
