The Infrastructure of Physical AI.
Investment Opportunity
While late-stage capital crowds into massive robot manufacturers (e.g., Humanoids), the early-stage alpha lies in the infrastructure layer: the edge perception, Vision-Language-Action (VLA) models, and agentic cybersecurity that makes physical AI possible. We are targeting companies providing the "Picks and Shovels" for the Embodied AI gold rush.
Investment Goals
Purpose: Strategic equity acquisition for high-growth venture returns, specifically seeking 10x-100x Multiple on Invested Capital (MOIC).
- Targeting Series A entries ($10M-$15M checks).
- Building a clustered portfolio around AI infrastructure and cyber.
- Identifying founders with deep technical moats.
"The next frontier of AI is physical AI. AI is now beginning to understand the laws of physics... The ChatGPT moment for general robotics is just around the corner."
— Jensen Huang, CEO of NVIDIA
"Text was first, video is happening now, but the ultimate embodiment of intelligence will be systems that can take actions in the physical world."
— Sam Altman, CEO of OpenAI
"We are transitioning from systems that simply tell you things, to autonomous agents that actually execute tasks on your behalf in the real world."
— Satya Nadella, CEO of Microsoft
"The bottleneck for robotics is no longer the hardware. It is the intelligence layer. Foundation models are finally allowing machines to reason in unstructured environments."
— Pieter Abbeel, UC Berkeley & Co-founder of Covariant
Market, Macro & ESG
Evaluating the macro-level drivers forcing a structural shift away from capital-intensive hardware toward the high-margin software, simulation, and security infrastructure of Physical AI.
Market Analysis: The Infrastructure Pivot
The physical AI market is undergoing a structural shift mirroring the early cloud computing era. Rather than betting on capital-intensive hardware manufacturing (which inevitably faces commoditization and margin compression), the smart money is moving to the software, simulation, and security layers. Morgan Stanley's latest robotics research indicates that while hardware TAM scales, the ultimate value capture will reside in the intelligence and orchestration layers.
Macroeconomic Considerations
Labor Shortages: The Capgemini 2026 report cites 74% of enterprises are scaling Physical AI immediately due to chronic labor deficits (projected 2.1M manufacturing deficit in the US by 2030).
Onshoring & Supply Chain: Geopolitical tensions and post-pandemic policies (e.g., CHIPS Act impacts) are forcing domestic manufacturing. High-cost domestic labor necessitates rapid robotic automation to maintain margins.
🌎 ESG Considerations
Environmental
AI-orchestrated logistics reduce supply chain emissions by optimizing routes and eliminating empty warehouse transit space. Simulation-first training reduces physical e-waste.
Social
Directly addresses the "3D" jobs (Dull, Dirty, Dangerous). Automating kinetic workflows dramatically reduces workplace injuries in manufacturing and construction.
Governance
Investing in "AI Cyber" (like Target #3) ensures kinetic robots are tamper-proof, addressing board-level compliance and safety standards for AI agents.
Value Prop & Growth Potential (Series A)
Select a company to view its unique value proposition and long-term scalability.
Select a target company from the list to view the investment thesis and founder analysis.
VC Investment Framework
Financial Projections
Underwriting metrics for Series A targets in the Physical AI stack:
- Target Entry Valuation $40M - $80M (Pre)
- Current ARR Baseline $1M - $3M
- Required YoY Growth 3x (T2D3 model)
- Target Return Profile 10x - 20x MOIC
Risk Assessment
Primary risks and mitigations:
Hardware Dependency
Risk: Startups burning cash on custom robotics. Mitigation: Invest *only* in hardware-agnostic models (VLA/Software).
Kinetic Hallucination
Risk: AI models making erratic physical moves. Mitigation: Require strict "Simulation-to-Reality" validation and back cyber-defense platforms (Target #3).
Compute Scarcity
Risk: Cost of GPU inference at the edge. Mitigation: Focus on lightweight, specialized models over broad AGI wrappers.
Exit Strategy
Time horizon: 5-7 years. Realization scenarios:
Strategic M&A (Most Likely)
Big Tech (NVIDIA, Microsoft, Apple) lacking physical edge logic will acquire VLA "brains." Enterprise giants (SAP, Oracle) will acquire orchestration platforms. Case Study: Vayu Robotics was successfully acquired by Serve Robotics (NASDAQ: SERV) in August 2025 specifically for its AI foundation models, validating the high exit demand for VLA middleware.
IPO Window
For companies defining a new category (e.g., universal OS for robotics), targeting public markets by 2030-2031 as macroeconomic rates stabilize.
Alignment with Seligman Strategy
Why Seligman Ventures is positioned to win in Series A infrastructure.
Technical DNA
Seligman's IIT roots mean we can deeply evaluate edge-native models and sensor latency tech better than generalist funds. We underwrite true mathematical and technical moats, not just API wrappers.
Cybersecurity Portfolio
As physical robots scale, they become kinetic endpoints. Startups securing VLA model APIs fit seamlessly into our existing cyber thesis (e.g., Exaforce), protecting the AI agent economy.
Kellogg Network
Leveraging the Kellogg MBA network provides our Series A founders with early access to enterprise GTM strategies, helping highly technical engineers translate their platforms into scalable B2B sales.
References & Data Sources
Methodology and primary sources utilized to underwrite market sizing, macroeconomic trends, and Series A valuations for this thesis.
Market Sizing & TAM
Morgan Stanley Research (2026): "The Next Tech Cycle: Physical AI Infrastructure." Referenced for the ~$50B TAM projection for robotics middleware and orchestration.
Bessemer Cloud Index: Referenced for comparative SaaS/Software margins (~80%) versus hardware manufacturing (~25%).
Macro & Labor Economics
Bloomberg Intelligence: Demographic and labor tracking. Sourced the 2.1M projected US manufacturing labor deficit by 2030.
Capgemini Research Institute: "The Future of Supply Chain Automation (2026)." Sourced the 74% enterprise adoption rate driven directly by acute labor shortages.
Venture & M&A Data
Crunchbase & PitchBook (Q1/Q2 2026): Verified global VC funding capture and recent Series A valuations for Archetype AI ($35M) and Alias Robotics (€5M + EIC).
SEC Filings / TechCrunch: Verified the August 2025 acquisition of Vayu Robotics by Serve Robotics (NASDAQ: SERV) used in the Exit Strategy case study.
Executive Consensus
Public keynotes, shareholder letters, and press releases (2025-2026) confirming the VLA shift. Includes direct quotes from Jensen Huang (NVIDIA CES Keynote), Sam Altman (OpenAI), Satya Nadella (Microsoft), and Pieter Abbeel (Covariant/UC Berkeley).