Intel RealSense Spins Out From Intel With Depth Perception Installed Base and Unverified Market-Share Claims

RealSense, Inc. completed its separation from Intel in 2025, emerging as an independent stereo depth perception hardware supplier with $50M in disclosed funding, 130 employees, and a self-reported installed base spanning autonomous mobile robots, humanoid platforms, and access control systems. The Cupertino-based company now operates under license from Intel for its trademarks and certain IP — a dependency that defines both its near-term credibility and its primary structural risk.

---

Business Model and Strategic Position

RealSense operates as an asset-light perception hardware supplier, generating revenue through depth camera sales, OEM depth modules, and the proprietary D4 Vision Processor ASIC. The company targets four verticals: production robotics, access control, industrial automation, and healthcare. No revenue, margin, or customer concentration data has been disclosed publicly.

The dormakaba strategic investment — the highest-confidence signal in the company’s recent activity — opens a regulated, high-volume access control vertical. dormakaba operates across 130 countries with products deployed in commercial real estate, hospitality, and critical infrastructure. Anti-tailgating and presence detection represent concrete near-term use cases for stereo depth in physical security, a segment with measurable procurement cycles and established integrator channels.

The AVerMedia partnership, announced in early 2026, targets embedded vision distribution for Physical AI applications. Signal confidence is MODERATE — the strategic rationale is coherent, but revenue contribution and deployment scope remain undisclosed.

---

Signal Activity — Intel RealSense

Competitive Positioning — Intel RealSense

Technology and Product Portfolio

Product	Interface	Enclosure	Target Application	Status
D457 Depth Camera	GMSL/FAKRA	IP65	Vehicle-grade, outdoor robotics	Fielded
D555 Depth Camera	PoE on-chip	Not disclosed	Enterprise fleet, fixed automation	Fielded
Depth Modules	Board-level OEM	N/A	Embedded integration	Fielded
Vision Processor D4/ASIC	Board-level OEM	N/A	Cost-optimized stereo depth	Fielded
SDK 2.0	Open-source	N/A	Developer/production integration	Fielded

The D457 and D555 represent a deliberate shift from developer-grade hardware toward production-qualified SKUs. The D457’s GMSL/FAKRA interface and IP65 rating directly address the objections that historically limited RealSense deployment in outdoor and vehicle-mounted applications — long cable runs, EMI exposure, and environmental ingress. The D555’s integrated PoE simplifies cabling architecture for fixed-installation fleets, a practical advantage in warehouse and facility automation at scale.

The D4 ASIC paired with board-level depth modules provides OEM customers a cost-optimized integration path without full custom sensor design. This creates meaningful switching costs: robot manufacturers that design around the D4 face non-trivial re-qualification costs to substitute an alternative depth solution.

SDK 2.0 remains the company’s most durable ecosystem asset. Its adoption across ROS, robotics research institutions, and production AMR stacks represents years of accumulated integration work that competitors cannot replicate quickly. GitHub and Discord community activity sustains developer familiarity that translates directly into design-in decisions at OEM level.

---

Market Position

RealSense claims 60% share of the global AMR market and 80% of humanoid robotics deployments. Both figures are self-reported with no independent third-party corroboration — LOW CONFIDENCE for investment purposes. CES 2026 ecosystem showcases with Unitree, LimX Dynamics, Mobile Industrial Robots, and Boston Dynamics via Intel Foundry provide directional support, but named deployments at disclosed volume would be required to underwrite the category-leadership narrative.

Competitive substitution risk is real. Time-of-flight sensors from Sony DepthSense, LiDAR miniaturization, and integrated sensor-fusion SoCs from well-capitalized competitors all address overlapping use cases. Stereo depth’s cost and power advantages in structured indoor environments remain intact, but the modality is not universally superior across all robotics deployment contexts.

---

Outlook and Key Catalysts

RealSense’s near-term credibility depends on converting inherited ecosystem momentum into independently verifiable commercial traction. Three specific catalysts would materially change the investment case: publicly announced multi-year OEM contracts with named AMR or humanoid manufacturers at disclosed volume; dormakaba deployments scaled to quantifiable unit counts; and independent market research corroborating the AMR and humanoid share claims.

The Intel IP licensing dependency remains the most significant unresolved structural risk. License terms, duration, exclusivity, and pricing are undisclosed. A change in Intel’s licensing posture — whether through renegotiation, restriction, or trademark enforcement — could constrain product roadmap freedom and margin structure simultaneously.

At 130 employees and $50M in funding, RealSense faces a resource allocation challenge across four verticals. Prioritization discipline will determine whether the company builds durable positions in robotics and access control or spreads too thin across healthcare and industrial automation before achieving profitability. The next funding round, if disclosed with valuation data, will be the clearest signal of independent viability.