The podcast features a discussion among industry experts in manufacturing, focusing on the evolution of technology, the impact of Ignition, and the collaborative nature of the industry. The conversation delves into the challenges faced by end users, the importance of community, and the need for education and guidance in adopting new technologies. The conversation delves into the challenges and opportunities of enterprise-scale digital transformation, focusing on the concept of Unified Namespace (UNS) and its impact on manufacturing and industrial operations. The discussion emphasizes the importance of intentional deployment, data management, and the role of technology in driving value and transformation. The conversation covers the strategic relationship between Fuuz and Inductive, the role of data model-driven solutions, the importance of standards, the value of AI in factory connectivity, and the responsible use of AI and UNS. It also explores the impact of MQTT and Sparkplug in transforming architectures and the significance of leveraging AI and UNS for customers.

Takeaways

• Industry 4.0 and digital transformation
• Community collaboration and open standards Enterprise-scale digital transformation requires intentional deployment and a focus on driving value through technology.
• The concept of Unified Namespace (UNS) plays a crucial role in creating interoperability and data management in industrial operations. Strategic relationship between Fuuz and Inductive
• Role of data model-driven solutions
• Importance of standards and AI in factory connectivity
• Responsible use of AI and UNS
• Impact of MQTT and Sparkplug in transforming architectures

Chapters

• 00:00 Introduction to the Full Stack of Manufacturing
• 05:48 Legacy Pain Points and Modern Tools
• 13:15 The New Paradigm in Manufacturing
• 20:18 The Role of Fuuz in Enterprise Layer
• 28:18 Challenges of Enterprise-Scale Deployment
• 33:18 Unified Namespace (UNS) and Interoperability
• 43:29 Success Stories and Adoption of UNS
• 52:08 Mindset and Approach to Technology
• 57:59 Value of AI and UNS for Customers
• 01:15:14 Impact of MQTT and Sparkplug in Transforming Architectures

Summary Dylan DuFresne and Zack Scriven go tool-by-tool through the five platforms that run most serious industrial stacks — Ignition, Flow Software, Litmus Edge, HighByte, and Fuuz — and answer the question every IT/OT team is asking: when do I use which, and why. This is Episode 2 of the Prove It series on Abelara Ascent, building toward the live reference architecture at the Prove It conference.

About the Speakers Dylan DuFresne leads architecture and digital transformation engagements at Abelara. Zack Scriven hosts Abelara's podcast and livestream content. Both are preparing Abelara's Prove It booth — a full Enterprise B reference architecture running locally on-site at the conference.

Key Topics

• The "motor overheating" problem that explains what data ops actually is
• Each platform's native home turf — composable SCADA, analytics, edge, integration, cloud-native MES
• HighByte vs Flow: data in motion vs data at rest
• The Prove It Enterprise B architecture — 3 sites, 5 platforms, different combinations per site
• Fuuz security: single outbound tunnel, no third-party surface
• Rapid-fire: when NOT to use each platform
• Pricing reality across all five vendors
• SESME / SMIPS: the interoperability initiative for shared data models

Key Takeaways

1. Start with a problem, not a technology. Every digital transformation that skips this tends to fail.
2. Data in motion → HighByte. Data at rest → Flow. Device connectivity at scale → Litmus. Plant-floor visibility → Ignition. Cloud-native enterprise backbone → Fuuz.
3. For small companies without process maturity, a clipboard and Excel beat any of these tools. People, process, technology — in that order.

Notable Quotes

"HighByte is for data in motion and Flow is best for data at rest." — Dylan DuFresne

"For the really small companies, a good process is way more important than the tools you're using." — Dylan DuFresne

"It's people, process, and technology in that order for a reason." — Dylan DuFresne

Timestamps

• [00:00] — Welcome + Prove It series recap
• [04:44] — The five platforms and their native homes
• [15:01] — Why more platforms doesn't mean more data models
• [18:35] — HighByte vs Flow: data in motion vs at rest
• [23:19] — The Prove It Enterprise B architecture walkthrough
• [28:26] — Fuuz security: the single outbound tunnel
• [37:24] — Rapid fire: when NOT to use each platform
• [44:23] — Pricing reality check
• [48:54] — SESME / SMIPS and the interoperability vision

00:00 — Intro + the planned stack

05:11 — The credibility moment: "All 8 title sponsors, by coincidence"

07:45 — Edge devices, Opto 22, and why IPCs run Linux

13:00 — "Edge is shorthand for as far into the stack as I care to look"

15:16 — The Ignition 8.3 tag server

18:02 — TimeBase vs. TimescaleDB, MES core vs. MES custom

24:51 — SCADA in Vision, MES in Perspective — why

27:31 — Medallion architecture, HighByte, and the enterprise UNS

34:34 — Fuuz as the enterprise Swiss Army knife

45:09 — Q&A

Summary

Zack Scriven sits down for an impromptu conversation with Dylan DuFresne (Lead Architect & Co-founder, Abelara) on three of the industry's most-abused terms: "digital transformation," "unified namespace," and "Industry 4.0." Dylan's core argument is that the terms themselves are buzzwords — vessels that vendors fill with whatever they're trying to sell. The only useful question is what problem are you actually solving, and the only useful follow-up is what is this vendor actually selling me.

The conversation moves from meta-critique into practical territory in the second half: Dylan walks through what a UNS has actually meant for the industry over the last decade (short answer: education), whether greenfield and brownfield sites need different answers (yes), and which tools he'd pick today if he were starting from scratch. It closes with one of the cleanest vendor-selection frameworks you'll hear on a podcast — and a discovery-call CTA for anyone who wants to talk through their own architecture.

Three takeaways

1. "Digital transformation" is whatever the vendor needs it to be to close the sale. The term has meant ERP-to-cloud migration, workflow digitization, knowledge-base digitization, and now unified namespace. The throughline isn't technical — it's commercial. If you're buying "digital transformation," demand to know what product or service is actually being delivered, and whether you have a problem it solves.
2. UNS is a tool, not an answer. The real value of the UNS movement was education. Over ~10 years, the UNS conversation did something no one argued against: it educated the market on the need to connect ERP to plant floor, and share data across layers that never shared it. But "I want a UNS" is still the wrong starting question. Greenfield deserves a UNS-shaped architecture. Brownfield with working historians usually does not.
3. Pick tools by problem, pick vendors by partnership. Dylan's ideal greenfield stack: Ignition, HiveMQ, Flow Software, Solace — some combination gets you most of what you need. Broker market has commoditized; the differentiation is adjacent features and culture. When two vendors solve the same problem, pick the one who wants to grow with you — not the one with the nicer salesperson.

Chapters
00:00 — Cold open + intro 

00:32 — What "digital transformation" actually means (and who benefits from the ambiguity)

03:56 — Why digital transformation is change management, not a product 

05:08 — Unified namespace: what it is, what it became, what it's worth 

08:30 — Greenfield vs. brownfield — and the real value of UNS 

14:56 — Who actually benefited from the UNS movement 

17:27 — The stack Dylan would build today (+ CoreFlux deep dive) 

19:24 — How to pick vendors when products overlap 

24:41 — Discovery calls + weekly architecture Q&A

Glenn Gardner (Abelara co-founder) sat down with Craig Scott (Fuuz CEO) at Fuuz HQ after spending a week embedded in bootcamp with the Fuuz team. The conversation threads through three topics: the unusual culture Fuuz has built around genuinely caring about manufacturing, the architectural decision to be the MES "shim" between ERP and the plant floor, and how democratized plant-floor data changes work for the personas nobody usually builds for — finance, accounting, and maintenance.

Craig Scott — CEO and founder of Fuuz, a manufacturing operations platform purpose-built to unify ERP, plant floor, and operational workflows in a single data layer. Background spans CNC and robot programming, owning a manufacturing shop, and running a systems integration practice before Fuuz.

1. Culture is a hiring choice, not a slogan. Fuuz didn't hire developers who already knew manufacturing. They hired strong developers and put them on real implementation projects, forcing them to learn manufacturing by solving actual customer problems. Their CTO Lance now runs a maker space on the side.
2. ERP and plant floor are different data models — stop trying to merge them. ERP is relational, transactional, 300 transactions/day. Plant floor is scalar time series, thousands of data points per second. Predictive maintenance is vector matrix tensor. You can't force all three through one pipe. What's needed is a unified data layer (the "shim") that speaks all three natively.
3. The forgotten personas are the highest-leverage users. Finance and accounting teams burn cognitive cycles trying to reconstruct variances with spreadsheets. Maintenance teams can't get their predictive data to the enterprise. A platform that democratizes plant-floor data to those personas is worth more than another BI dashboard.

"I found the park of people like me." — Glenn Gardner, on the Fuuz team's culture

"ERP doesn't have to be those things, but there has to be that shim, that MES layer in there." — Craig Scott

"I still have ulcers from those periods." — Glenn Gardner, on quarter-close variance hunting

• 00:00 — Cold open: MES theory vs. reality
• 00:38 — The Bumblebee culture metaphor
• 03:37 — Why Fuuz hired devs and taught them manufacturing
• 09:26 — What kind of SI Fuuz was, and how Fuuz emerged
• 13:00 — The MES shim explained (ISA-95 / Purdue layers)
• 17:11 — Why iPaaS approaches fail
• 19:10 — Scalar vs. time-series vs. tensor data
• 21:57 — 300 ERP transactions vs. millions of telemetry events
• 24:08 — The AWS outage and Fuuz's architectural resilience
• 25:11 — Predictive maintenance and the vector-tensor database problem
• 28:25 — Democratizing data to finance and accounting
• 31:19 — The "ulcers" story — quarter-close variance chasing

• Fuuz: https://www.fuuz.com
• Abelara: https://www.abelara.com
• Craig Scott on LinkedIn: https://www.linkedin.com/in/craigascott1/

Summary

The Abelara team — Glenn Gardner, Zack Scriven, and Dylan DuFresne — just spent a week at Fuuz corporate headquarters for bootcamp. In this roundtable debrief, they give their unfiltered take: what they loved, what needs work, how Fuuz compares to Ignition and Plex, where it fits in the stack, and why a greenfield site should probably run both Ignition and Fuuz together. The conversation covers real pricing comparisons, the MCP tooling announcement, and why "it does everything" is both Fuuz's greatest strength and its biggest go-to-market challenge.

Key Topics Covered

• What each team member is most excited about after bootcamp
• What Fuuz needs to improve: UX, user journeys, discoverability of features
• Fuuz vs. Ignition: not competitors — different layers of the stack
• Fuuz vs. Plex: why enterprises evaluating Plex should look at Fuuz
• Pricing reality: single-site MES is comparable to Ignition; multi-site is where Fuuz pulls ahead
• Edge gateway architecture: cloud platform with an on-prem edge component
• MCP tooling: Fuuz is building it now while others have only announced it
• The three pillars: schema designer, flow designer, screen designer
• Why "it does everything" is both the value prop and the sales problem
• Build vs. buy spectrum: Fuuz as the middle ground between full custom and rigid off-the-shelf

Key Takeaways

1. Ignition owns the lower stack (SCADA, plant floor). Fuuz owns the upper stack (enterprise MES, iPaaS, ERP integration). A greenfield site should run both — they're best friends, not competitors.
2. Fuuz's hardest problem isn't the product — it's explaining the product. It's six or seven commercializable products in one platform, and nobody can pitch that in a sentence without sounding like BS.
3. MCP tooling on the Fuuz flow designer means a lot

Notable Quotes

• "Ignition and Fuuz should probably be the best of friends." — Dylan DuFresne
• "It's like seven different applications. And if someone said pitch me on Fuuz, I would sound like an idiot." — Glenn Gardner
• "Other leading platforms have announced the ability to do this in the future. Fuuz is doing it now." — Dylan DuFresne (on MCP)

Timestamps / Chapters

[00:00] — What stood out most from Fuuz bootcamp
[01:56] — Zack: enterprise orchestration and best-of-breed integration
[03:08] — Dylan: the people, the platform, and where Fuuz fits in the stack
[04:48] — Honest critique: what needs to improve
[06:48] — The discoverability problem and the change logging trap
[09:31] — Edge gateway explained: cloud platform with edge component
[13:02] — "Fuuz is like six or seven different products"
[15:12] — Pricing reality: Fuuz vs. Ignition vs. enterprise iPaaS
[17:47] — Single-site vs. multi-site: where Fuuz takes the lead [20:06] — Ignition and Fuuz: complementary, not competitive
[21:44] — Fuuz vs. Plex: entirely different platforms
[23:29] — What module would you add next? SCP.
[24:40] — The most common problem: integrating ERP with the plant floor
[32:55] — MCP tooling: the AI unlock
[33:41] — Glenn's final take: card-carrying Fuuz fan

Links & Resources

• Fuuz: fuuz.com
• Abelara: abelara.com
• Ignition: inductiveautomation.com

Summary

Glenn Gardner frames the quality problem every manufacturer faces: DPPM spikes and nobody can figure out whether it's a catastrophe or a blip — because the data lives in a dozen different systems. Steve Modrall walks through how Fuuz solved this for HighBar, the world's first solar-powered steel mill, deploying MES, WMS, yard management, scale house, and vendor portal in nine months. Craig Scott demos Fuuz's schema designer and shows how embedded quality checks and digital work instructions eliminate the "more training" corrective action that plagues every 8D process.

Key Topics Covered

• Quality as the hardest problem in manufacturing: tracking DPPM across downstream, midstream, and upstream
• Why ERP falls down for midstream and upstream quality data
• HighBar Steel: world's first solar-powered steel mill, 2x output per FTE vs. industry
• Red data vs. blue data: enterprise governance vs. plant-level detail
• iPaaS as the bridge between ERP and plant floor (not point-to-point)
• WIP visibility without SKU proliferation
• Schema designer: modeling manufacturing data as normalized relationships
• Poke-yoke through MES: embedding quality into process, not as an afterthought
• 8D root cause failures when you run out of data
• MCP integration for AI/LLM analysis on manufacturing data

Key Takeaways

1. The ERP handles downstream data (RMAs, customer returns) well, but for midstream (factory floor) and upstream (supplier quality), you need MES, SCADA, and quality systems — the ERP was never designed for that level of detail.
2. Don't track WIP in the ERP — it explodes your SKU count by an order of magnitude. Keep raw materials and finished goods in ERP; let the MES own the WIP detail.
3. If your 8D corrective action says "more training," you've run out of data. Embed quality checks into the process so the failure literally cannot propagate forward.

Notable Quotes

• "Your business doesn't happen in the ERP. It happens on the plant floor." — Zack Scriven (quoting Walker Reynolds)
• "You start making stuff up because you're out of data." — Glenn Gardner (on 8D failures)
• "They're running this mill with half the staff and generating double the output per FTE." — Craig Scott (on HighBar)

Timestamps / Chapters

[00:16] — Welcome and introductions
[01:34] — Glenn: the manufacturing reality check
[03:02] — Quality as the most direct impact to revenue, margin, and customer sat
[05:14] — Downstream data: ERP handles RMAs well
[06:30] — Midstream: where ERP falls down
[08:20] — Upstream: supplier quality
[12:08] — "Your business happens on the plant floor" [13:22] — Steve: HighBar Steel case study
[15:35] — Fuuz iPaaS: agnostic integration vs. point-to-point [24:08] — Craig: MCP and AI/LLM on manufacturing data [27:47] — Why serial-level detail doesn't belong in the ERP [30:28] — Fuuz as the unified namespace for this architecture
[34:34] — WIP in ERP = SKU hell
[36:21] — Red data vs. blue data explained
[40:45] — Craig: live demo of the schema designer
[54:00] — Poke-yoke through MES
[57:09] — ISO 9001: embedded procedures vs. printed SOPs
[59:29] — The drunk guy looking for his keys: why 8Ds fail

Links & Resources

• Fuuz: fuuz.com
• Abelara: abelara.com
• Oracle NetSuite: netsuite.com

Summary

Glenn from Abelara visits Fuuz corporate headquarters during boot camp week and sits down with Steve to walk through the real manufacturing pain points that Fuse solves — from the supply chain and factory floor divide, to quality traceability, bill of materials wars, real-time variance tracking, and the multi-plant standardization trap. Steve frames Fuse as the system that lives "where the carpet meets the concrete."

About the Guest

Steve is a senior leader at Fuuz with deep manufacturing operations experience. He positions Fuse as a composable platform covering MES, WMS, quality, scheduling, EDI, and iPaaS integration — designed to complement ERP rather than replace it.

Key Topics Covered

• The supply chain vs. factory floor divide: pegging sales orders to work orders
• ERP's visibility gap below the shop floor level
• Quality traceability: tracking components from dock to shipment
• Recording PLC parameters per serial number for root cause
• The bill of materials war: supply chain, manufacturing, and R&D views
• The golden triangle: PLM ↔ MES ↔ ERP integration
• Real-time variance tracking by work center and operator
• Multi-plant standardization without forcing ERP onto the floor
• Reducing tech stack from 40 vendors with a composable platform
• Balancing corporate control vs. local plant innovation

Key Takeaways

1. Pegging sales orders to work orders lets planners make partial shipment decisions in real time — most systems produce blanket work orders blind to customer demand.
2. The million-unit recall problem exists because companies test everywhere but trace nowhere — traceability is the missing layer, not more testing.
3. Use ERP for financials and front office; use MES for the shop floor. Forcing ERP down to the plant floor is why no enterprise has ever achieved one-ERP standardization.

Notable Quotes

• "Where the carpet meets the concrete — that's where Fuse takes over." — Steve
• "The money's made and lost on the shop floor." — Steve
• "No two machines are alike. Even from the same manufacturer, they're all different." — Steve

Timestamps / Chapters

[00:00] — Inside Fuse HQ: a software company that knows manufacturing
[01:34] — The supply chain and factory floor divide
[03:14] — Pegging sales orders to work orders
[04:30] — ERP's visibility gap on the shop floor
[06:55] — Quality: the hardest challenge in manufacturing [08:09] — Tracking components from dock to shipment [09:48] — The million-unit recall problem
[11:05] — Recording 120 PLC parameters per serial number [12:48] — What R&D leaders really need from production data [14:18] — The bill of materials war
[15:29] — The golden triangle: PLM, MES, and ERP
[18:53] — Real-time gross margin and variance tracking [21:42] — IT's challenge: future-proofing manufacturing [24:04] — The multi-plant standardization trap
[27:49] — Employee development across standardized plants
[29:48] — Closing thoughts

Links & Resources

• Fuse: fuse.com
• Steve Modrall: LinkedIn
• Glenn Gardner: LinkedIn
• Abelara: abelara.com

Dylan DuFresne walks through Abelara’s reference architecture for modern industrial systems — from edge controllers through Ignition’s tag gateway pattern, UNS publishing, historian integration with Timebase, Flow Software for contextualization, and enterprise-scale deployment with HiveMQ cloud bridging and Fuuz as an MES layer. The episode covers how to phase these components in with zero technical debt and how Abelara helps enterprises cut through analysis paralysis.

Dylan DuFresne is an architect at Abelara specializing in Ignition, MQTT / Unified Namespace, and enterprise data systems.

• Tag gateway pattern: separating tag providers from front-end / back-end DevOps
• Ignition 8.3 improvements and Gateway Area Network updates
• Historian integration: Timebase direct from tag gateway vs collectors
• Flow Software for contextualization, KPIs, and bidirectional flow
• Blue namespace vs red namespace: site flexibility vs enterprise standards
• HiveMQ cloud bridging to Snowflake, BigQuery, Azure
• Fuuz as an MES layer replacing ETL between ERP and site systems
• Phased implementation: SCADA → UNS → historian → analytics → AI

• Use a dedicated tag gateway so front-end and back-end systems can run DevOps cycles without risking production tag state.
• Standards should get stricter as you move up the stack — fluid near controllers, rigid near ERP.
• Build additively with zero technical debt: SCADA, then UNS, historian, analytics, AI.

[00:00] Tag gateway pattern
[02:42] Timebase historian
[04:01] Flow Software
[07:45] HiveMQ + cloud data lakes
[13:04] Fuuz as enterprise MES
[15:06] Zero technical debt roadmap
[17:50] Abelara workshops

Abelara: https://abelara.com
Ignition: https://inductiveautomation.com
Flow Software: https://flowsoftware.com
Timebase: https://timebased.io
HiveMQ: https://hivemq.com
Fuuz: https://fuuz.com
Litmus: https://litmus.io
Axilon: https://axilon.com

Dan Prudhoe spent 15 years building industrial systems at a large chemical manufacturer — scaling from beta Ignition in 2009 to 70+ gateways across 30+ sites. In this ICC 2025 conversation, he walks through replacing a legacy Unix MES with Ignition, adopting Spark Plug B and MQTT, hitting MQTT's limitations, and building an enterprise event mesh with Solace that streams millions of tags from edge to cloud.

About the Guest

Dan Prudhoe is a Senior Solutions Engineer at Solace. Previously he spent 15 years at a large chemical manufacturer leading Ignition adoption from beta, building a standardized platform across 30+ sites, and architecting an event mesh for millions of historized tags.

Key Topics Covered

• Ignition from beta (2009) to enterprise-scale across 30+ plants
• Building modular, upgradeable standards across diverse sites
• Spark Plug B and MQTT: decoupling systems, lightweight data collection
• MQTT limitations for transactional and guaranteed delivery
• Event mesh with Solace: edge to cloud, multi-protocol, guaranteed messaging
• Self-service analytics with Seeq
• UNS debate: centralized semantic layer vs. composable hierarchies
• Graph databases and industrial data modeling
• The path: self-service → automated → autonomous (agentic AI)

Key Takeaways

1. Start with analytics, not SCADA — lower risk, proves value fast, builds political capital for mission-critical systems later.
2. Build standards designed to flex — the first few deployments will break your standard, and that's expected.
3. The path to autonomous manufacturing: self-service (SMEs solve their own problems), automated (event-driven workflows), then autonomous (agentic AI investigates and prescribes).

Notable Quotes

• "We wanted to build a standard that's meant to have flexible standards." — Dan Prudhoe
• "Instead of we don't have enough data, they're almost like oh my god there's so much data." — Dan Prudhoe
• "We were able to allow them to start solving their own problems." — Dan Prudhoe

Timestamps / Chapters
[00:00] — Who is Dan Prudhoe?
[01:06] — Ignition beta testing in 2009
[02:20] — MES 101
[05:10] — Scaling to 70+ gateways across 30 sites
[08:53] — In-house vs. system integrator
[11:01] — Ignition becoming enterprise-grade
[12:00] — From Spark Plug B to event mesh
[15:14] — MQTT's limitations
[16:51] — Solace: edge to cloud architecture
[20:55] — Self-service analytics for SMEs
[26:31] — Predictive maintenance data provisioning
[28:30] — The UNS debate
[32:08] — Graph databases and ontology
[33:42] — Wrap-up

Links & Resources

• Solace: solace.com
• Dan Prudhoe LinkedIn
• Ignition: inductiveautomation.com
• Flow Software: www.flow-software.com
• Spark Plug B: sparkplug.eclipse.org
• Azure Data Explorer: azure.microsoft.com/en-us/products/data-explorer