Safety Third: Autonomous Revolution

Reynolds and Moore

Welcome to our podcast dedicated to the intersection of innovation and safety. Join us as we engage with subject matter experts across various industries, exploring the pivotal query of ensuring safety innovation. We delve into conversations with safety industry experts to glean diverse perspectives, invaluable experiences, and groundbreaking concepts in the realm of safe innovation. Your host, Erik Reynolds, guides these discussions, offering a platform for insightful dialogue and strategic insights.

Play

Open in

22 episodes
fortnightly
Avg 54 min
English

Episodes 22

Newest Oldest Longest

All20 Unplayed20 In progress0 Played0

Counted on this page — what you have heard stays on this device, so it is not something the list can be paged by.

Monday · 50 min
Building the Safety Hardware Underneath Robots with Weihan Wang
Weihan Wang, Director at NexCOBOT, a NEXCOM International company, joins Erik Reynolds to talk about the hardware underneath robot safety. Weihan leads business development, product development, and functional safety for the robotics division and is the safety manager for the SCB 100. They cover how a motion controller, a PLC, and a safety controller differ, and what makes a safety rated PC a different product from a regular one. Weihan shares how NexCOBOT spent 4 years taking the SCB 100 through SIL 2 and PL d Category 3 certification. He also explains why force monitoring pushes robot makers away from safety PLCs and toward C and C++ on a PC platform, and how customers run redundant safety applications on 2 cores inside a single chip. The conversation then turns to humanoids, quadrupeds, and ISO 25785, which the ISO working group expects to publish around 2028. Weihan also describes the safety platform NexCOBOT is building on NVIDIA IGX Thor, where a safety island monitors the AI inference and a separate safety MCU serves as the backup. R&M and NexCOBOT are sharing a booth at NVIDIA GTC Berlin on October 20–22. Stop by to meet both teams. More on NexCOBOT: https://www.nexcobot.com/ Check out our work: https://reynolds-moore.com/
September 7 · 50 min
When Code Drifts From Its Architecture with Dr. Daniel Simon
Dr. Daniel Simon joins Erik on the podcast to talk about why the structure of safety critical software decides how safely it can be built and changed. Daniel is Distinguished Solutions Engineer at Qt Group, a member of the R&M Safety Partner Ecosystem. He co-originated Axivion Architecture Verification out of research at the University of Stuttgart and stayed with the work after Qt Group acquired Axivion. They cover what architecture verification checks, and why documented architectures go stale a few years into a project. Daniel explains why moving compute intensive code from CPU to GPU is an architectural exercise rather than a refactor. He also walks through the Axivion work on CUDA, including the collaboration with NVIDIA on their safety coding rules. The conversation ends on AI generated code and what it does to the person who has to sign off on it. Daniel is also speaking at the R&M Physical AI & Safety Summit on October 13 in Boston at MassRobotics, where he will cover why deterministic verification and emerging AI trust methods under IEC 61508 and ISO/IEC 22440 have to coexist as robotics teams pursue certification. The Summit is free to attend, and an evening reception at Qt Group follows. The next 2 days at MassRobotics, Reynolds & Moore is running Understanding ISO/IEC TS 22440, a training in collaboration with SafetyCert.AI. Register for the Summit: https://wkf.ms/4cdJRlF Register for the 2-Day Training: https://wkf.ms/4gBG3Mj More on Qt Group: https://www.qt.io/ Check out our work: https://reynolds-moore.com/
August 24 · 48 min
The Race Between Standards and Technology with Chris Soranno
Chris Soranno, Manager of International Safety Standards at SICK Sensor Intelligence, joins Erik to take on a complaint the industry makes constantly, that safety standards are always behind the technology. Chris has spent 25 years inside that process. A technology has to prove it is not a passing trend before a committee will invest the effort, and by the time the first meeting happens, the market has already moved again. His argument is that a standard should say what has to be achieved, never how to achieve it. Chris points to the safety laser scanner, a proven technology that had nowhere to be applied because the application standards of the day named the light curtain as the only accepted answer. They also cover why risk assessment still holds up for machines whose behavior is harder to predict, why motion and contact are no longer treated as automatically hazardous, and what changes when the thing defeating your safeguard is an adversary rather than an accident. More on SICK Sensor Intelligence: https://www.sick.com/ Check out our work: https://reynolds-moore.com/
August 9 · 54 min
The New European Rulebook for Intelligent Machines with Lorenzo Nava
Erik talks with Lorenzo Nava, European Managing Director of Reynolds and Moore, about the new regulatory landscape facing intelligent machines in Europe. They look at how the Machinery Regulation, the Cyber Resilience Act, and the AI Act overlap, and what that means for manufacturers trying to bring AI-driven machinery to the European market. Lorenzo explains the shift from prescriptive standards toward more flexible, framework-based ones, and why defining a clear operational design domain matters when certifying complex systems like humanoids. They discuss the current state of machine learning certification, the standards still in development, and whether notified bodies are ready to sign off on these technologies. The conversation also covers market surveillance obligations across physical, cyber, and AI systems, and the tensions that arise when requirements to record events run up against European privacy rules. Check out our work: https://reynolds-moore.com
July 20 · 49 min
The Deliberate Path to Humanoid Safety with Kevin Reese
This week, Erik Reynolds talks with Kevin Reese, Distinguished Robotics Safety Engineer at Agility Robotics and project leader for ISO 25785-1, the first safety standard for dynamically stable industrial mobile robots. They discuss why classical machine safety assumptions break down for legged robots, where a standard stop can be the most hazardous event, and how the new DSIMR standard fits alongside existing robot and industrial truck standards. Kevin shares how Agility is expanding Digit's capabilities intentionally, starting with cooperative operation in industrial settings before taking on harder problems like collaborative work and unstructured environments. They also talk about the role of AI in safe human detection, what ISO/IEC TS 22440 means for using AI in safety functions, and what 65,000 operating hours across nine customer deployments has taught the team that the lab never could. Check out Agility Robotics: https://www.agilityrobotics.com/ Check out our work: https://reynolds-moore.com/
July 6 · 1 hr 9 min
Inside MassRobotics: Where Robots Get Real with Russell Nickerson
Erik Reynolds sits down with Russell Nickerson, Partnership Engagement Liaison at MassRobotics, the world's largest independent nonprofit robotics hub, and founder of Collabots, an entertainment robotics consultancy with clients including Cirque du Soleil. Russell shares what he sees every day working with startups, corporate partners, and academia: the gap between a robot that works in the lab and one that survives the real world. They dig into UX failures, E-stop recovery on a live stage, why demo day is when robots find new ways to fail, and the overlooked infrastructure problem nobody is talking about, the shortage of integrators and repair shops that can keep robots running long after deployment. They close on what separates the startups that reach deployment from the ones that don't, and why Russell believes the answer has less to do with the robot and more to do with the customer. Join Reynolds & Moore at MassRobotics in Boston this October: R&M Physical AI & Safety Summit, October 13, 1:00-5:00 PM EDT. Free to attend. Register here. Understanding ISO/IEC TS 22440: Artificial Intelligence – Functional Safety and AI Systems, a two-day in-person training in partnership with RMAI, October 14-15. Seats are limited. Register here. Check out MassRobotics: https://www.massrobotics.org/ Check out Collabots: https://collabots.net/ Check out our work: https://reynolds-moore.com
June 21 · 1 hr 5 min
The Virtual Operator with Phillip Hansel
Phillip Hansel, CEO and Co-founder of AIOps, joins Erik Reynolds to talk about what happens when reinforcement learning shows up on the factory floor. Phillip spent more than a decade as a control systems engineer in oil, gas, and chemical processing, and now builds AI for the same kind of plants he once worked in. They cover the virtual operator, an AI agent trained to run a process the way a seasoned human would. They get into why shaping that agent's behavior makes the modern controls engineer as much psychologist as engineer. The conversation also gets into the plants where these agents already run, a 20-year control problem one of them quietly solved, and why Phillip is adamant this technology exists to back up human operators rather than replace them. More on ⁠AIOps: https://ai-ops.com/ More on Reynolds & Moore: https://reynolds-moore.com/
June 8 · 49 min
From a Blank Page: Writing the Rules for AI Safety with Jordan Punch
Jordan Punch, Senior Functional Safety Engineer at Reynolds and Moore, joins Erik Reynolds to talk about what it takes to write a safety standard for AI from scratch. As a voting member of ISO SC42 JWG4, the working group developing ISO/IEC TS 22440, Jordan has spent the past two years helping shape the first international standard for AI functional safety. They cover the committee draft process and the roughly 1,700 comments national bodies returned, each of which has to be answered before publication. They get into the difference between a technical report, a technical specification, and a full international standard, and what each means for certification. The conversation also gets into the publication timeline, why it could slip from early 2027 into later that year, and the one thing Jordan wants engineers and safety managers to take away as 22440 moves toward release. Want to understand ISO/IEC TS 22440 in more depth? Reynolds and Moore is hosting two upcoming events for you: ISO/IEC TS 22440: Inside the Standard, a 45-minute webinar with Jordan on June 30, 2026 at 10 AM CDT. Understanding ISO/IEC TS 22440, a two-day training in Boston on October 14–15, 2026. Register Today: www.reynolds-moore.com/events Check out our work: www.reynolds-moore.com
May 25 · 56 min
One Year Later: Sensory Robotics & the Speed of Safety With Chris Edwards & Mark Gagas
Erik welcomes back Sensory Robotics' Chris Edwards, Founder and CEO, and Mark Gagas, COO, one year after their first appearance on Safety Third: Autonomous Revolution. In the year since their first appearance, Sensory Robotics completed verification and validation of the SR1 system under UL 1740 after 19 months of work, navigating standards that were not written for their technology. The system now includes an AI body tracking layer, with the data feeding into a new offering called SR Insight. The conversation covers the certification process, their go-to-market approach through system integrators, the DoD SBIR project they secured, and where humanoid robotics fits into the safety picture. More on Sensory Robotics: https://sensoryrobotics.com Check out our work: https://reynolds-moore.com/ Missed Chris and Mark's first episode? Listen here.
May 11 · 51 min
The Safety Functions That Don’t Exist Yet with Dr. Ricard Picas Prat
In this episode, Erik Reynolds talks with Dr. Ricard Picas Prat, Advanced Solution Engineer at Novanta Inc., about functional safety for dynamically stable robots. Ricard brings a background spanning academic research, hardware engineering, and years of leading SIL 3 and PL E functional safety certifications at what is now Novanta's servo drives division. They cover how classical safety functions like safe torque off and safe brake control work, and why those approaches fall apart for humanoids and quadrupeds. When a dynamically stable robot loses power, it doesn't stop safely. It falls. The conversation also gets into the challenge of writing a standard for technology that is still being invented. Ricard is doing that work firsthand as an active contributor to ISO TC 299 Working Group 12, while simultaneously building products for the market those standards will eventually govern. More on Novanta Inc: www.novanta.com Check out our work: www.reynolds-moore.com
April 27 · 1 hr 12 min
How Reynolds & Moore Was Built: 10 years of Growth, Risk, & Functional Safety with Erik Reynolds & Missy (Moore) Reynolds
In honor of Reynolds & Moore's 10 year anniversary, co-founders Erik and Missy Reynolds sit down with producer Ginny Spellman to talk about how the company got built and how the functional safety industry has changed over that time. They cover the founding story, the culture Erik and Missy wanted to build from day one, and what it took to grow from a one-person consultancy to a global team. The conversation also gets into how the market for functional safety has shifted over the last decade, the role Missy is stepping into around developing and training Reynolds & Moore engineers, and where the company is headed in the next 10 years. If you are interested in becoming a living kidney donor, visit nkr.org/GRT523. Check out our work: www.reynolds-moore.com
April 13 · 53 min
The Final Frontier: Machine Learning Safety in Robotics With Paul Schmitt
Paul Schmitt, Director of Engineering at Reynolds & Moore, joins Erik Reynolds to explore what the safety community still has not solved: how robots decide what to do. Paul brings decades of experience across Ford, iRobot, and autonomous trucking to a conversation about the behavior layer in physical AI. They cover human-robot interaction research, including a study on how pedestrians actually respond to driverless vehicles, and why the answer had nothing to do with lights or sounds. Paul and Erik also dig into behavior hierarchies and the trolley problem applied to real robotics scenarios. The conversation also gets into why current machine learning safety standards still read more like guidelines than engineering requirements. Paul shares details on his work as a co-author of "The ML FMEA in Action: Lessons from Applications of Machine Learning Safety," documenting how the open-sourced ML FMEA framework connects machine learning development decisions to system-level safety outcomes, with applications across autonomous vehicles, humanoid robotics, and beyond. Read Paul's paper: The ML FMEA in Action: Lessons from Applications of Machine Learning Safety Check out our work: www.reynolds-moore.com
March 30 · 1 hr 5 min
What Certification Bodies See That Engineers Don't with Bodo Seifert
Bodo Seifert, Principal Automotive Functional Safety Engineer and Practice Lead at TÜV Rheinland, joins Erik Reynolds to share what certification bodies actually see when companies come in for assessment. They cover the organizational and technical red flags that assessors look for, and why good engineering practice and a formal safety case are not the same thing. Bodo draws on experience across automotive, robotics, and autonomous trucking to explain what separates companies that certify successfully from those that do not. The conversation also gets into the shift from deterministic functional safety to AI-enabled systems. Bodo explains what ISO/TR 5469 and ISO/IEC TS 22440 mean for teams building physical AI, and why transparency and explainability remain the hardest requirements for machine learning developers to accept. They also discuss why strategic collaborations like the one between Reynolds & Moore and TÜV Rheinland matter for companies navigating certification for the first time. Check out our work: www.reynolds-moore.com
March 16 · 50 min
Engineering Trust with Dr. Werner Kraus
In this episode of the Safety Third Podcast, Erik Reynolds speaks with Dr. Werner Krauss, head of the research division for automation and robotics at the Fraunhofer Institute for Manufacturing Engineering and Automation (IPA), where he has spent 15 years bridging academic research and real-world industrial deployment. They cover the real engineering challenges behind deploying humanoid robots in industrial settings, drawing on Fraunhofer IPA's hands-on evaluation of the UNITREE G1 across safety, cybersecurity, clean room performance, and what those findings mean for trusting these platforms in the real world. Werner also shares what surprised his team most about these platforms, why end users are harder to design for than the robots themselves, and the incident from his own career that changed how Fraunhofer trains everyone who operates a robot. Read the Fraunhofer IPA's 3-part series on humanoid robots in industry. Check out our work: www.reynolds-moore.com
March 2 · 47 min
One Platform for Functional Safety with Josiah Rohne
Erik welcomes back Josiah Rohne, Staff Systems Engineer at Fennec Engineering, to discuss ASAP, Fennec's full-lifecycle functional safety platform, and the shift from fragmented tools to a single connected workflow. ASAP connects risk assessment, requirements, design, testing, and the safety case through shared traceability, while Fennec Engineering also develops dedicated test environments for safety validation, with the results feeding back into the platform to support certification. The conversation looks at scaling safety beyond a single owner, managing the impact of hardware and BOM changes, and why closing the gap between development and real-world validation is becoming critical. Josiah closes with a reflection on why the real measure of success in safety is often the incidents that never happen. Fennec Engineering is part of the R&M Safety Partner Ecosystem. Reynolds & Moore is hosting a 45-minute webinar in collaboration with Fennec Engineering, Modernizing Functional Safety: Building Your Safety Case Without Spreadsheets on March 31, 10:00 AM - 10:45 CST. Register here. Missed Josiah's first episode? Listen here. Check out our work: https://reynolds-moore.com/
February 16 · 1 hr 7 min
Scaling Safety Innovation with Cameron Gieda
Erik sits down with Cameron Gieda, Executive and Business Development leader at FORT Robotics, to discuss the commercial strategy required to move safety technology from lab prototypes to real-world deployment. They explore the stakeholder stack, explaining why the success of an innovation often depends more on aligning IT, finance, and operations than on the engineering itself. Cameron breaks down the massive economic disconnect between the $700 billion spent on insurance and liability versus the relatively small investment in preventive technology. The conversation also covers the rise of humanoid robotics and why collaborative partnerships like the R&M Safety Partner Ecosystem are essential for small companies to deliver enterprise-scale solutions. Cameron closes with a high-stakes story about an autonomous vehicle demo gone wrong that illustrates the thin line between breakthrough and disaster. More on FORT Robotics: https://www.fortrobotics.com/ Check out our work: https://reynolds-moore.com/
February 2 · 48 min
Inside the AI Safety and Inspection Lab with Anita Dodia
In this episode of Safety Third: Autonomous Revolution, Erik talks with Anita Dodia, Laboratory Manager at RMAI, about how testing, inspection, and certification are evolving for AI-enabled safety systems. They explore why traditional functional safety standards struggle with black box AI models, and what it really means to validate systems that are probabilistic, non-transparent, and shaped by complex real-world environments. Anita explains how robustness testing, simulation, and physical validation can be used to define safe operating boundaries for physical AI. The conversation also touches on emerging standards like ISO/IEC TS 22440, along with the role of accreditation and collaboration across safety and standards community. They discuss why integrating safety into AI system design early on, rather than treating it as a late-stage requirement, determines whether autonomous systems can realistically be tested, validated, and certified. More on RMAI: www.safetycert.ai Check out our work: www.reynolds-moore.com
January 19 · 49 min
The Cultural Divide Between AI, Safety, and Security with Patrik Feth
Erik sits down with Patrik Feth, Professor of Computer Science at International University of Applied Sciences (IU) and AI Officer at NewTec, to discuss what it actually takes to move artificial intelligence from research into real world safety critical systems. The conversation explores the “valley of death” in technology deployment and the need to bridge the gap between traditional functional safety and modern software security. Patrik explains why the industry must move past “safety by obscurity” and the “not invented here” mindset to embrace a more collaborative, open culture. They also explore the shift from deterministic to probabilistic engineering and the cultural changes required to build truly certifiable and trustworthy autonomous systems. Check out our work: www.reynolds-moore.com
January 5 · 49 min
Sensing Through Walls with Natalya Lopareva
Natalya Lopareva, co-founder and CEO of Algorized, joins Erik to talk about non-visual human sensing and why true perception goes beyond cameras for the future of robotics. From detecting breathing and heartbeats through walls to creating a real-time safety bubble around people in industrial environments, this conversation explores how machines can understand human presence through a head to head ground truth verification process. The discussion touches on Edge AI, latency, and the transition from rescue missions to commercial robotics. Natalya also reflects on how her approach to risk and innovation was shaped by her experience racing across the Pacific Ocean, where she navigated extreme weather conditions and experienced the surreal shift of living the same day twice after crossing the time zone. More on Algorized: https://www.algorized.com/ Check out our work: https://reynolds-moore.com/
Dec 22, 2025 · 1 hr 4 min
Zero is a Real Number with Roberta Nelson Shea
In this episode, Erik Reynolds sits down with Roberta Nelson Shea, Global Technical Compliance Officer at Teradyne Robotics, covering Universal Robots and MiR, and ISO convener for TC 299, for a wide-ranging conversation on how safety evolves across robotics, automation, and emerging technologies. Drawing on decades of experience and her role in shaping standards such as ISO TS 15066 and ISO 10218, Roberta discusses why modern safety standards focus on goals rather than prescriptions, where compliance and safety diverge, and why machines cannot interpret risk the way humans do. The conversation spans collaborative robots, legacy factory equipment, autonomous vehicles, and AI-enabled safety systems, including a candid exchange on expectations around risk and the idea that “zero” is still a number when it comes to safety outcomes. Check out our work: https://reynolds-moore.com/

Showing 1–20 of 22 episodes