About Rose City Robotics
Our Mission
CAD-driven defect detection for precision manufacturers.
At Rose City Robotics, we build inspection from CAD so quality is defined up front and stays stable on the line. Developed by Joseph Cole, PhD, RoseVision is built for quality and manufacturing engineering teams running high-mix production — shops where visual inspection is impossible to standardize, escaped defects are a real operational risk, and you can't wait months to collect training data before your next revision ships.
We go beyond thresholds and templates. Upload your STEP file, define the defects that matter, and deploy reliable go/no-go inspection on on-prem edge hardware — no cloud, no fragile rule engines, no full retraining when your CAD changes. Joseph Cole brings 20 years of machine vision and factory automation experience to every engagement.
Our Team
We’re machine vision engineers, automation specialists, and manufacturing technologists who’ve worked in real factories — not just research labs. We partner directly with quality engineers and process engineers who run production: helping design pilots, wiring into robotic cells, and generating the data that makes the business case. We don’t hand over software and disappear. Led by Joseph Cole, PhD, we bring 20 years of industrial vision and automation experience to every engagement — and we stay until it works.
-
Joseph Cole, Ph.D.
Co-founder, CEO / Technology and Vision
Joseph earned his PhD in applied physics from Rice University and a graduate certificate in applied statistics from Portland State University (PSU). He is a retired Major with the US Army Reserves and has over 20 years of experience developing computer vision and machine learning algorithms across a variety of industries.
-
Duncan Miller, MBA
Co-founder, Operations / Business Development
Duncan is a software engineer with over 20 years of experience as an entrepreneur. He is a FIRST Robotics coach and has a passion for technology and entrepreneurship. Duncan earned an MBA from Babson College and is a mentor and innovation competition judge at the PSU Center for Entrepreneurship and Business Accelerator.
-
Aaryn Curl
Senior Program Manager Robotics & AI
Aaryn is a lifelong engineer with 19 years at Intel, where she led cross-functional software programs from concept to launch. At RCR, she combines her technical experience and passion to enable our customers to get the most out of their software and hardware investments.
-
Kaj Hansteen Izora
Associate Physical AI Engineer
Kaj (pronounced like Kai) is an engineer in AI, ML, and robotics, specializing in simulation, interpretability, and applying novel statistical methods to machine learning. He is an expert in utilizing tools like Omniverse and Cosmos to build realistic simulations of physical systems.
-
Samantha Town
Marketing and Communications
Sam helps shape our communication and operational strategies with a conscientious emphasis on human relationships and impact. Sam graduated from the University of Oregon and grew up working in a family business in Oregon wine country.
Our Advisory Board
Our distinguished advisors bring deep technical expertise and real-world experience. They provide strategic counsel on technical direction, partner ecosystems, and deployment realities so we can scale CAD-driven inspection in high-mix environments without sacrificing reliability. Their guidance helps us deliver practical, software-first quality control that teams can trust from day one.
Our Values
We believe factory automation should be honest about what it can do, built to last in real production, and delivered by engineers who stand behind their work.
- Integrity
-
We operate with transparency, not just talk. We're self-capitalized — no VC, no PE, no investor clock ticking. That means every decision we make is filtered through one question: does this serve our customers? We don't ask for recurring fees until you have hard data that the system works in your environment.
- Innovation
-
We build from your CAD, not from scratch. From STEP-to-model toolchains to on-prem Jetson deployment, we iterate on real factory constraints — and we show up on the floor to make it work. We're reading research papers and watching BattleBots so you get battle-tested innovation, not vendor hype.
- Impact
-
We measure success at the end of the line. Fewer escaped defects. Less over-processing. Unattended runs that actually run unattended. When a quality engineer can walk their VP through a business case built on real pilot data — that's the outcome we're after.
Machine Vision Lab
Our team of PhDs and engineers help precision manufacturing and automation teams turn vision and imaging ideas into production-trustworthy algorithms. We don’t build cells or sell cameras — we plug in as the algorithm team alongside your existing engineering and integration partners.
Recent work includes image registration from in-procedure video in surgery, and creating high-fidelity virtual worlds for training physical AI systems (synthetic environments and simulation pipelines that let teams validate perception + inspection logic before full-scale deployment).
- Vision & Imaging Algorithm Development
- Design and implement robust vision/imaging algorithms (classical + ML) for inspection, defect detection, measurement, tracking, and image registration — built to survive real-world variability.
- Feasibility & Risk Evaluation
- Quickly assess whether a vision/imaging use case is viable with your real samples and constraints. Clear answers on expected performance, key risks, and what it would take to get to production reliability.
- Reliability & Performance Improvement
- When a system “works in the demo” but not in production: diagnose failure modes, reduce false rejects/misses, and tighten evaluation so improvements are measurable and repeatable.
- Edge / On-Device Optimization
- Optimize algorithms for real-time, embedded, or edge constraints (latency, throughput, memory). Make models and pipelines run where they need to run — reliably (Jetson-class or x86).
- Synthetic Environments & Simulation
- Build high-fidelity virtual worlds and simulation pipelines to validate perception + inspection logic before full-scale deployment. Useful for autonomy, robotics, and complex environments where data collection is slow or expensive.
- Industrial Inspection Systems (On-Prem)
- Defect detection and measurement workflows for manufacturing — designed to hold up under production variance. We can work within strict data/security constraints and integrate outputs into your PLC/robot/HMI stack.
Our Company
Engineering-led. Software-first quality control.
Rose City Robotics is a Portland-based machine vision lab led by Joseph Cole, PhD. We build inspection from CAD so QA is defined up front and stays stable on the line.
RoseVision CAD Studio turns your design files into deployable inspection logic—ready for on-prem edge hardware like NVIDIA Jetson Orin or x86.
Watch a quick walkthrough of CAD‑driven inspection.