The Robot Report Podcast

Our guest this week is Eric Sivertson, VP of Security Business at Lattice.

This episode explores why physical AI and humanoids change the security conversation from traditional IT risk to real-world safety risk.

Sivertson explains how FPGAs can act as deterministic guardrails for robotic and cyber-physical systems.

We discuss why safe is not enough if a robot is also connected, how zero trust and cyber resilience apply on the factory floor, and why hardware needs to verify, monitor, and recover systems in real time.

He says traditional factory networks relied on “guns, guards, and gates,” and once OT systems connect to IT and the cloud, the attack surface expands dramatically, and cyber resilience becomes essential.

###

Register now for RoboBusiness 2026: https://cvent.me/w0eRN9?RefId=podcast

On the show today our guest is John Black, CTO of Brain Corp. John shares how Brain Corp has evolved from proving autonomous navigation in public spaces to operating at fleet scale across cleaning, scanning, retail analytics, and new manipulation use cases.

He explains why the real challenge is not just making robots smarter, but building the guardrails, infrastructure, and platform reuse needed to deploy them safely and reliably in the real world.

John joins Gene Demaitre and Mike Oitzman to unpack how Brain Corp moved from proving autonomous navigation in public spaces to operating fleet scales across cleaning, inventory intelligence, and shelf scanning.

He explains why the company’s next leap is less about building one perfect robot and more about creating a platform that can support many robot form factors, many applications, and many customers at once.

You’ll discover:

Why public spaces are the hardest test for autonomous robots, and how Brain Corp designs for safety without killing productivity

How Brain OS evolved into a full-stack platform with cloud infrastructure, OTA updates, manufacturing tools, trust centers, and fleet management

Why 50,000 connected robots and tens of millions of operating hours create a data advantage smaller fleets simply cannot match

If you want a clear-eyed look at where physical AI is headed — and what it takes to make robots commercially useful instead of just impressive — this episode delivers the strategy behind the scale.

###

Register now for RoboBusiness 2026: https://cvent.me/w0eRN9?RefId=podcast

Roby Lynn, founder and CEO of R2 Labs, shares how his team is bringing software-defined automation to industrial manufacturing.

He explains how the R2 Autonomy Controller connects PLCs, robots, vision systems, MES/ERP software, and other factory assets into one configurable platform.

The conversation explores the growing gap between legacy OT systems and modern IT tools, and how R2 Labs is helping manufacturers unify workflows, improve visibility, and add intelligence without replacing the systems they already trust.

Roby also reflects on lessons from building a company in a decades-old industry, including the importance of customer feedback, practical design, and solving real problems over chasing “cool” technology.

###

Register now for RoboBusiness 2026: https://www.robobusiness.com/

Design for manufacturing is changing faster than most robotics teams can keep up - and if you are building physical products, this conversation could save you months of rework. Marc Kermisch shares how Protolabs is using AI, simulation, and deep manufacturing expertise to turn CAD files into real parts in as little as 24 hours, while helping engineers avoid the design mistakes that quietly kill speed, quality, and scale.

Mike Oitzman and Gene Demaitre sit down with Marc, who has returned to the show since episode 138 with a new role as CTO and AI leader at Protolabs. He explains how the company’s software-driven approach links CAD models directly to manufacturing tool paths, G code, and production workflows across CNC machining, 3D printing, injection molding, and sheet metal - all built around the challenge of helping engineers get parts made faster without sacrificing precision.

You’ll discover:

- Why Protolabs treats AI as a practical manufacturing tool, not hype

- How machine learning helps catch manufacturability issues before a part is ever built

- What ProDesk does when it flags ejector locations, tight tolerances, seam issues, and other hidden design risks

- How part similarity search and simulation speed up internal decisions for engineers

- Where AI is already paying off in visual inspection, cobot programming, and print-box optimization

Marc also breaks down the real-world tradeoffs between 3D printing, CNC machining, and injection molding, including when a prototype should stay a prototype - and when it’s time to redesign for production. He gets specific about common failure points like draft angles, wall thickness, shrink, resin changes, tooling assumptions, and the gap between prototype tolerances and production reality.

If you’re dealing with robotics, hardware, manufacturing, or any physical product that must move from concept to production, this episode shows what happens when software, AI, and manufacturing expertise work together instead of in silos. It’s especially valuable for founders, roboticists, and engineers who need to make smarter decisions before the first expensive mistake happens.

Protolabs is also building for the future of compliance, supply chain resilience, and low-volume production, with a network that helps customers de-risk sourcing, reduce complexity, and stay aligned with regulated industries like defense, aerospace, and medical devices. The result is a rare inside look at how modern manufacturing is evolving - and how the next wave of physical products will get made.

Essential listening if you are building hardware, scaling production, or trying to make your robot, part, or process easier to manufacture the first time.

This episode explores how Nomagic is applying AI and robotics to warehouse operations, with a focus on each picking, recovery workflows, and production-grade deployment.
Josh Cloer, General Manager for North America, explains why the company leans into “physical AI,” how its systems are designed for always on operations, and why real-world production data matters more than simulation alone.
Mike Oitzman and Gene Demaitre also dig into the practical side of automation adoption, from pilot-to-production failures to the pressure on supply chain leaders to move faster without getting stuck in vendor hype.
The conversation is especially useful for teams evaluating warehouse robotics, AI-assisted recovery, or flexible automation strategies.
Learn more: https://nomagic.ai/