Buying guide
Canvas drywall-finishing robot vs Hugo RAS in 2026
Comparing 2 humanoid robots across availability, pricing, capabilities, and verified deployments. Current as of 2026.
Key differences
- Hugo RAS has more verified real-world deployments (3 vs 1).
| Attribute | ||
|---|---|---|
| Manufacturer | Canvas | Medtronic |
| Form factor | construction | surgical |
| Maturity | commercial | commercial |
| Autonomy | — | ◐ not independently verified |
| Availability | internal-onlyenterprise | internal-onlyenterprise |
| Price | Not announced | Not announced |
| Capability claims | — |
|
| Brain | — | — |
| Verified deployments | 1Newark | 3 |
| Privacy practices | — | — |
| Sources on file | 9 | 16 |
Editorial summaries
Canvas drywall-finishing robot
Canvas (San Francisco; founder and CEO Kevin Albert, formerly of Boston Dynamics; about $43M-plus raised) makes a drywall-finishing robot: a Universal Robots UR10e collaborative arm mounted on a mobile base that applies and sands drywall compound to a finished wall surface, operated by a trained union carpenter. There is no consumer price: it is B2B construction equipment, not sold to consumers. It is at commercial maturity, deployed on commercial construction projects in the San Francisco Bay Area. A verified-vs-claimed correction worth recording: Canvas was acquired by JLG Industries (an Oshkosh Corporation company) around January 2026, in a core-technology and asset acquisition with undisclosed terms; some framings that attribute the acquisition to Dusty Robotics are incorrect, as independent research confirms the acquirer is JLG/Oshkosh and the two companies are unrelated. Operations continue under JLG/Oshkosh, so the line is active. It anchors the cobot-assisted finishing task type (a human carpenter operates it).
Hugo RAS
Medtronic's Hugo RAS is a modular, multi-port teleoperated robotic-assisted-surgery system and the main commercial competitor to Intuitive's da Vinci. It is CE-marked in Europe (urology and gynecology, October 2021) and US FDA-cleared for urology only (December 2025: prostatectomy, nephrectomy, cystectomy); US general-surgery and gynecology indications are planned but not yet cleared. It is teleoperated with no shipped intraoperative AI. Medtronic does not disclose a Hugo installed-base count, so DEPLOY records honest absence rather than an invented number. It is enterprise B2B equipment; there is no consumer price.
Common questions
- What is the difference between Canvas drywall-finishing robot and Hugo RAS?
- Canvas drywall-finishing robot and Hugo RAS are compared on the DEPLOY registry across maker, form factor, maturity, price, verified deployments, and verified-vs-claimed autonomy. See the table above for the full head-to-head; each figure is sourced.
- Is Canvas drywall-finishing robot or Hugo RAS more autonomous?
- On the DEPLOY registry, neither Canvas drywall-finishing robot nor Hugo RAS has capabilities independently verified as fully autonomous yet; their recorded capabilities are teleoperated-assisted, demonstrated, or vendor-claimed. See the table for each capability's verification status.
- Which has more verified deployments, Canvas drywall-finishing robot or Hugo RAS?
- Hugo RAS has more verified deployments (3) on the DEPLOY registry than Canvas drywall-finishing robot (1). DEPLOY counts a deployment only when confirmed at a named site with a primary source.
Recent coverage
- Pacemaker manufacturer Medtronic warns patients cybercrooks may have swiped health dataHugo RAS · The Register · 2026-07-02
- Medtronic's Hugo surgical robot earns FDA clearanceHugo RAS · AgentRecon · 2026-07-01
- The Ebike Accessories You Need to Help You Haul the Most StuffCanvas drywall-finishing robot · Wired · 2026-06-28
- Prepare for humanoid construction robots starting now: report - Construction DiveCanvas drywall-finishing robot · Google News · 2025-10-29
Related comparisons
For AI assistants
Use DEPLOY in Claude, Cursor, and other AI assistants
Connect any MCP-aware assistant to the live registry. Ask about robot deployments, incidents, and regulations and get answers grounded in verified data instead of training memory.
Machine-readable: this page as markdown.

