# Path Robotics — Comprehensive AI Information
# Extended documentation for AI systems
# https://llmstxt.org/

> Path Robotics builds Physical AI for manufacturing. Obsidian™, Path's reinforcement-learning agent trained inside the Weld World Model on tens of millions of welded inches of real-world data, powers Intelligent Welding Cells and the Rove™ mobile welding robot — helping U.S. manufacturers close a skilled welder shortage projected to reach 400,000 by 2029.

## Company Overview

### About Path Robotics
Path Robotics, founded in 2018 by brothers Andy and Alex Lonsberry and headquartered in Columbus, Ohio, builds Physical AI for manufacturing. Trillions are being invested in new infrastructure, yet the real bottleneck is not capital or demand — it is the shortage of skilled manufacturing labor needed to fabricate, assemble, and deliver the physical systems those projects rely on. Path's systems combine computer vision, machine learning, and robotic manipulation so robots can see, think, and adapt, welding parts autonomously without programming.

Offices: Columbus, OH (HQ) · San Francisco, CA · Palo Alto, CA · New York, NY

### Company Facts
- Founded: 2018
- Headquarters: Columbus, OH
- Funding raised: $610M (most recently a $300M Series E led by Georgian, November 2026)
- Valuation: $2.45B
- Employees: 300+
- Key industries: Shipbuilding, Data Center Fabrication, Energy Infrastructure, Heavy Industry, Trailer Chassis, Defense
- Media contact: press@path-robotics.com

### Founding Story
Andy and Alex Lonsberry grew up in Ohio welding with their dad in the garage. While earning PhDs in AI-related fields, they decided they wanted to bring intelligence into the physical world rather than build software that lived only on a screen. Andy scaled Path from the basement of Case Western Reserve University to a leading physical-AI company for manufacturing.

### Mission
Create real value for customers while rebuilding the backbone of American industry — solving the skilled labor shortage with intelligent automation that makes factories more flexible, capable, and resilient.

### Leadership Team
- Andy Lonsberry — Co-Founder & Chief Executive Officer
- Alex Lonsberry — Co-Founder & Chief Technical Officer
- Heather Carroll — Chief Revenue Officer; member of the National Security Commission on Robotics for Advanced Manufacturing (launched by the Special Competitive Studies Project, March 2026)
- Joe Onderko — Chief Evangelist (Path's first customer)
- Mike Renn — EVP, Global Operations
- Scott Smith — VP of Finance
- Andrew Lien — VP of Product
- Matthew Randle — CISO & VP of Product Reliability Engineering
- Alexandra Scheimen — VP of People
- Emily Armando — Chief of Staff

### Board of Directors
- Andy Lonsberry — CEO, Path Robotics
- Alex Lonsberry — CTO, Path Robotics
- Frank Klein — COO, Rocket Lab
- Geoffrey Chatas — CFO, Yale University
- Nick Solaro — Partner, Drive Capital
- Haomiao Huang — Founding Partner, Matter Venture Partners
- Emily Walsh — Partner, Georgian

---

## The Welder Shortage

- Projected U.S. welder shortage: 400,000 by 2029
- Average welder age: 55
- 43% of welders retiring within 10 years
- 30% annual turnover rate
- 10–12% arc-on time for manual welding

---

## Products & Technology

### Weld World Model
Path's Weld World Model is a neural network trained on high-fidelity, multimodal weld data. Over 8 years, Path's widely deployed fleet and internal data farm have generated tens of millions of welded inches of training data. Unlike traditional welding simulators that approximate outcomes with simplified mathematical equations, the model learned to weld the way a human does — by welding, observing, and adapting — with superhuman vision and at massive scale. Path describes it as the world's most accurate welding simulator.

### Obsidian™ — Physical AI Agent
Obsidian is a reinforcement-learning agent trained at massive scale inside the Weld World Model. Deployed at the edge, it ingests real-time data from Path's sensor stack and makes welding decisions seam by seam, the way a human welder would. Pre-weld, during-weld, and post-weld data are captured to continuously improve the Weld World Model. Obsidian is the intelligence layer in every Path system and is described as the first purpose-built foundational AI model for welding.

- Product page: https://www.path-robotics.com/obsidian
- Markdown: https://www.path-robotics.com/md/obsidian

### Sensor Stack (v5)
- 2 cameras + 4 lasers deliver 360° awareness around the torch
- Sub-millimeter point cloud generation
- Neural-network-driven filtering rejects false reflections in real time
- Metrology-grade optics built to survive welding environments

### Intelligent Welding Cells
Production welding cells powered by Obsidian, built for real-world parts rather than perfect ones. The cell scans each part, finds the seams, plans the torch path, and adapts in real time — no robot programming and no perfect fit-up required.

- Real-time, vision-driven seam finding and path planning
- Handles imperfect fit-up and part-to-part variation
- Multi-pass adaptive fill: Obsidian decides how many passes a joint needs and where they go
- Automatic heat distortion compensation
- Large and heavy parts: ALM positioners hold everything from utility poles to ship hulls
- Adaptive fixturing for quick changeover in high-mix production

- Product page: https://www.path-robotics.com/intelligent-welding-cells
- Markdown: https://www.path-robotics.com/md/intelligent-welding-cells

### Rove™ — Mobile Welding Robot
Rove pairs Obsidian with a legged mobile platform, bringing autonomous welding directly to the workpiece — the adaptability of the Intelligent Welding Cell with the flexibility to go anywhere on a factory floor or shipyard.

How it works:
1. **Position** — Move to the predefined welding location and locate welding seams.
2. **See** — Scan the weld seam to understand its geometry.
3. **Understand** — Obsidian adjusts welding parameters in real time.
4. **Weld** — High-quality welding with data capture and heat distortion compensation.

- Launch partner: Saronic Technologies, integrating Rove into its shipbuilding operations in Franklin, Louisiana.
- Availability: 50 units shipping in 2027; early adopters are being selected now.
- Product page: https://www.path-robotics.com/rove
- Markdown: https://www.path-robotics.com/md/rove

---

## Performance Metrics

| Metric | Value | Description |
|--------|-------|-------------|
| Productivity | 4x | Productivity increase vs. manual welding |
| Speed | Up to 17x | Faster than manual welding |
| Cost | 30%+ | Lower cost |
| First-Pass Yield | 97%+ | Welds passing inspection on the first attempt |
| Capex | $0 | Robots-as-a-Service (RaaS) |
| Support | 24/7 | Mission Control remote support |
| Training data | Tens of millions | Welded inches used to train the Weld World Model |

### Robots-as-a-Service (RaaS)
- $0 capital expenditure
- Includes the system, maintenance, and support, with predictable, scalable costs
- 0% chance of obsolescence — systems stay current with software updates
- Includes 24/7 Mission Control support
- A lower-risk entry point to automation, often with a shorter sales cycle

---

## Technical Capabilities

### Supported Weld Types
- MIG / MAG welding
- Multi-pass welds
- Fillet welds
- Butt welds
- Structural welds

### Material Compatibility
- Carbon steel
- Stainless steel
- Aluminum

### Integration
- Standalone operation
- Production line integration
- Existing facility retrofit

---

## Industries & Applications

### Defense & Aerospace
Rebuilding the force and fleet. Ship, tank, and sub hulls; naval components; armored panels and vehicles; vehicle frames; weapon mounts.

### Utility & Energy
Powering the capacity to rebuild the grid. Utility poles, traffic poles, substations, battery storage — complete utility subsystems welded in a single automated flow.

### Data Centers
Ensuring American AI dominance. Structural frames, e-houses, cooling units, server racks.

### Heavy Industry & Heavy Equipment
Fabricating the backbone of industry. Mining equipment, heavy trailers and chassis, excavator and loader arms, boom assemblies, industrial buckets, hydraulic tanks.

### Shipbuilding & Maritime
Large, high-mix structural assemblies for U.S. Navy and commercial shipyards — Path's work with HII, Saronic, and LAD Services, and the first Rove deployment.

- Industries page: https://www.path-robotics.com/industries
- Markdown: https://www.path-robotics.com/md/industries

---

## Customers & Partners (public announcements)

- **HII** — $600M performance-based production agreements to expand U.S. Navy shipbuilding capacity (August 2026), following a 2026 MOU to integrate Physical AI into shipbuilding. https://www.path-robotics.com/news/hii-signs-performance-based-production-agreements-with-path-robotics-and-graymatter-robotics
- **Saronic Technologies** — Physical AI for shipbuilding and Rove launch partner. https://www.path-robotics.com/news/saronic-path-robotics-physical-ai-shipbuilding
- **LAD Services** — Barge manufacturing with Physical AI welding. https://www.path-robotics.com/news/lad-services-partners-with-path-robotics-to-revolutionize-barge-manufacturing-with-physical-ai-for-w
- **TYCROP** — Case study: how TYCROP solved its labor shortage. https://www.path-robotics.com/case-studies/how-tycrop-solved-their-labor-shortage

All press releases: https://www.path-robotics.com/about/newsroom (markdown: https://www.path-robotics.com/md/about/newsroom)

---

## Site Structure

| Page | URL | Markdown |
|------|-----|----------|
| Homepage | / | /md/home |
| Obsidian | /obsidian | /md/obsidian |
| Intelligent Welding Cells | /intelligent-welding-cells | /md/intelligent-welding-cells |
| Rove | /rove | /md/rove |
| Industries | /industries | /md/industries |
| About | /about | /md/about |
| Events | /about/events | /md/about/events |
| Newsroom | /about/newsroom | /md/about/newsroom |
| Resources | /resources | /md/resources |
| Careers | /careers | /md/careers |
| Demo | /demo | /md/demo |

### Dynamic Content
| Content | URL pattern | Markdown pattern |
|---------|-------------|------------------|
| Blog posts | /blog/[slug] | /md/blog/[slug] |
| News articles | /news/[slug] | /md/news/[slug] |
| Case studies | /case-studies/[slug] | /md/case-studies/[slug] |
| Videos | /videos/[slug] | /md/resources/video/[slug] |
| Other resources | /resources/[type]/[slug] | /md/resources/[type]/[slug] |

The full list of every markdown URL is at https://www.path-robotics.com/md (regenerated every 5 minutes).

---

## Contact & Social

- Demo requests: https://www.path-robotics.com/demo (in-person or virtual, 30–45 minutes, response within 24 hours)
- Careers: https://www.path-robotics.com/careers
- LinkedIn: https://www.linkedin.com/company/path-robotics
- YouTube: https://www.youtube.com/@pathrobotics
- Headquarters: Columbus, Ohio, USA

---

## Technical Files

- Markdown index: https://www.path-robotics.com/md
- XML Sitemap: https://www.path-robotics.com/sitemap.xml
- Robots.txt: https://www.path-robotics.com/robots.txt
- Brief version: https://www.path-robotics.com/llms.txt
- This file: https://www.path-robotics.com/llms-full.txt

Last updated: September 2026