What I Learned at IoT Tech Expo Global 2026
- Yoel Frischoff
- 2 days ago
- 5 min read
By Yoel Frischoff | February 2026 | TheRoad
I spent two days at Olympia London last week, wandering the halls of IoT Tech Expo Global 2026. After 25 years in product strategy – much of it at the intersection of hardware and software – I've developed a filter for these events: ignore the buzzwords, watch for the shifts in conversation.
This year, the conversation shifted.
Why London, Not Vegas
CES gets the headlines. The glitzy product reveals, the celebrity appearances, the "world's first" announcements that may or may not ship. It's a consumer spectacle, and it serves that purpose well.
But if you're a European corporate trying to figure out your IoT strategy, or a vendor selling into enterprise and industrial markets, CES isn't where the real conversations happen.
IoT Tech Expo fills a different niche. The audience skews toward practitioners – engineering leads, operations directors, procurement teams actually evaluating solutions for deployment. The exhibitors aren't chasing consumer buzz; they're showing connectivity modules, edge platforms, security architectures, and integration services. The conversations are less "wouldn't it be cool if" and more "how do we make this work at scale."
What makes the London event particularly valuable is the co-location with AI & Big Data Expo. For anyone working on smart products, this combination is increasingly essential – IoT provides the data streams, AI provides the intelligence layer. Walking between halls, you see the full stack: sensors and connectivity on one floor, machine learning platforms and analytics on the other. The convergence isn't theoretical here. It's happening in adjacent booths.
For European companies especially, there's another dimension: regulatory alignment. The EU's evolving stance on data sovereignty, cybersecurity requirements, and sustainability reporting means that European vendors and buyers need to talk to each other directly. An event in London – still a gateway between European and global markets – offers that proximity.
This year, the conversation shifted.
"Smart" Finally Means Something
For years, "smart" was just a marketing suffix. Slap it on a thermostat, a lock, a refrigerator – instant (quasi) differentiation, no substance required. The word became so diluted it almost lost meaning.
What I noticed at this expo is that the conversation has matured. The vendors who matter aren't selling "smart" as a feature anymore. They're selling relationships.
Take Glean, which was present at the co-located AI & Big Data Expo. Their enterprise AI platform doesn't just search your company's data – it learns, connects, and delivers increasingly personalized answers over time. The value compounds with use. That's not a product sale; it's an ongoing relationship where the platform gets smarter the longer you stay.
That's the model smart tangible companies should be studying. Not "we sold you a device," but "we're continuously delivering value through what we learn." The initial sale is just the handshake. The real asset is the relationship that follows.
Edge AI Has Quietly Arrived
Three years ago, edge AI was aspirational. Vendors showed demos that worked on stage but struggled in production. The gap between "AI-capable chip" and "deployed AI solution" was a chasm.
That gap is closing fast.
Edge Impulse, Analog Devices, Microchip – they're all showing production-ready solutions for MCU-class devices. Not prototypes. Not roadmaps. Shipping products.
Fiona Treacy from Analog Devices gave a talk on running machine learning models directly on edge devices. Hugh Breslin from Microchip tackled what he called "building ultra-reliable intelligence at the constrained IoT edge." The emphasis on reliability tells you something: we've moved past proving it works to ensuring it works consistently.
My prediction? Embedded intelligence becomes table stakes by 2027. If your connected product doesn't have some local processing capability, you'll be competing on price alone.
The Digital Twin Grows Up
Daniel Isaacs from the Digital Twin Consortium presented a framework that stopped me in my tracks. He described the evolution of digital twins in three stages:
Traditional twins handle visualization – a 3D model of your asset.
Intelligent twins add analytics – the model learns from operational data.
Generative twins create scenarios – the AI generates simulations and recommendations.
That third stage is new. We're no longer just monitoring physical assets through their digital representations. We're using AI to imagine futures for them. The digital twin becomes a decision-making partner, not just a dashboard.
For anyone building smart products, this has profound implications. The twin isn't just a service add-on. It's becoming the primary interface between manufacturer and deployed product – long after the sale.
Security: From Feature to Market Access
Cyber-physical security is a years long concern. This expo add solutions to this critical problem.
Mohamed Abdelmawla from NXP ran a fireside chat on "designing connected devices with security built in." Not bolted on. Built in. The language of secure-by-design has moved from security conferences to mainstream IoT events.
More telling: exhibitors like ONEKEY are building businesses around compliance management for IoT products. When there's a market for managing your security certifications, you know the procurement conversation has changed.
Security is no longer a feature. It's becoming a market access requirement.
The People Who Made It Worthwhile
Expos are about conversations as much as presentations. I met Antony Fryer from OpenText, who's been thinking deeply about enterprise IoT use cases – he has that rare combination of business and technical fluency that makes for genuine insight.
Alex Dzenisionak, CEO of Streamlogic, caught my attention with his framing of organizations that "compound value, not just ship features." That's exactly the platform mindset I've been documenting in my book work.
I connected with several Israeli colleagues – Yoav Crombie from AGAT Software AI and Neal Harow from Deepchecks – which reminded me how concentrated the IoT security and AI validation communities have become in our little corner of the Mediterranean.
Mark Waters from CENSIS (Scotland's innovation centre for sensing, imaging, and IoT) brought a practitioner's perspective on edge ML deployment. These are the people actually wrestling with production constraints, not just theorizing about them.
What I'm Taking Home
Three convictions reinforced by this expo:
The relationship is the product. We're past the era of selling hardware and hoping customers come back. The companies winning now are building ongoing relationships through software, services, and data. The initial sale is just the handshake.
Local intelligence is non-negotiable. Cloud-dependent products are becoming liabilities—for latency, for privacy, for reliability. The edge isn't just an optimization. It's becoming the expectation.
Security is a strategic asset. Not a cost center. Not a compliance checkbox. The companies that build trust through demonstrable security will command premium positioning as regulations tighten and buyers get smarter.
The expo hall is packed up now. The banners are rolled away. But the conversations keep going – on LinkedIn, in follow-up emails, in the ideas that make it into product roadmaps.
That's what makes these events worthwhile. Not the swag. The signal.
Let's Talk
If you're navigating similar challenges – figuring out how to layer software and services onto physical products, or wondering what your business model looks like when the hardware is just the beginning – I'd welcome a conversation.
IoT Product Strategy Consulting I work with hardware companies and IoT ventures on the strategic questions that don't have obvious answers: What services justify recurring revenue? How do you architect digital layers that create defensible value? Where does the business model actually live?
AI Augmentation for Connected Products The edge AI shift I described above isn't theoretical—it's a design decision you're making right now. I help teams think through AI deployment across IoT architectures: swarm intelligence, command and control systems, and the analytics split between edge and cloud. The goal is augmentation that compounds value, not complexity.
If either resonates, reach out.
Yoel Frischoff is a product strategist and consultant based in Tel Aviv, with 25+ years of experience across healthcare, education, consumer goods, and IoT. He runs TheRoad, a boutique practice focused on software-augmented hardware strategy.