Patrick J Huston | IoT × Supply Chain × Security

Visibility is not safety. Intelligence is. The supply chain industry spent a decade and billions of dollars learning to see everything. It forgot to learn what to do about it. A thread on the most expensive mistake in modern logistics →

Studying executive strategy at Berkeley taught me something uncomfortable: most companies don't have a strategy problem. They have a translation problem. The vision exists. The KPIs exist. What's missing is the 90-day action that connects them. Goals without short-term execution milestones are just corporate poetry.

I’m a 90 year old man trapped in a 40 year old body. The beautiful @NFTMrsPeach took me to The French Laundry to celebrate the occasion. A truly once in a lifetime experience

6.9% vacancy in refrigerated warehousing is a real estate story. The story underneath it is an operations story. When companies pre-position temperature-sensitive inventory across surplus facilities to hedge tariff risk, you get more product sitting in more locations with less visibility per site. Nobody's asking who's monitoring condition compliance across all that newly distributed inventory. The square footage is available. The intelligence layer isn't.

"Self-optimizing factories: 50% productivity gains with IoT + edge AI + robotics." Translation: those numbers come from greenfield showrooms built to prove the concept works. The other 90% of facilities are brownfield. Legacy PLCs that speak 4 different protocols. Sensors bolted onto equipment older than the engineers maintaining it. "Integration" means a guy with a clipboard walking between two systems that don't talk to each other. I've walked into 200+ facilities across 60 countries. The gap between the conference keynote and the loading dock is where most digital transformation projects go to die. The real innovation isn't the factory of the future. It's making intelligence work inside the factory of 1987.

Unpopular opinion: reshoring doesn't reduce supply chain risk. It relocates it. Domestic supply chains have better surveillance. Shorter lead times. Familiar regulatory environments. They also concentrate operational risk into fewer nodes. One port backlog, one rail disruption, one labor action can cascade faster when you don't have geographic diversification absorbing the shock. I've watched companies exit Asia and move production stateside, then discover their new domestic network has 3 single points of failure they never had before. The threat model shifts. It doesn't disappear.

A distribution company ran into this wall so many times they built a tech company to fix it. The 3PL pre-sales process hasn't changed since the 90s. Sophisticated WMS inside, cold calls and outdated directories outside. The search tools are evolving faster than the 3PLs marketing through them. Same pattern in cargo: we invested billions in tracking what happens in transit. Almost nothing in verifying the carrier before the shipment moves. The unsexy discovery layers are where the real gaps live.

Digital twins keep failing and nobody wants to say why. It's not the model. It's the gap between the simulation and the warehouse floor. Employees don't trust the twin because the twin doesn't reflect what they see every morning. Stale data in, confident predictions out. thescxchange.com/tech-infrastru…

@antgrasso The same logic applies to cargo. Every shipment already carries ownership, handling rules, and custody transfers across networks. The difference is most of that information still lives in PDFs and phone calls instead of programmable infrastructure.

Tokenization turns assets into programmable units carrying ownership and rules across networks. Markets open to smaller tickets and new investors, so operations move with fewer intermediaries and clearer audit trails, pushing firms to adapt capital flows. Microblog @antgrasso

@OFOEAC What percentage of containers actually get any physical inspection before clearing? K9 teams are effective but the math doesn't scale to volume. That's where sensor-layer screening starts to matter.

From aircraft to cargo containers, CBP K9 teams detect narcotics, currency, and other contraband across air and maritime environments. Trained, certified, and mission-ready to support officers wherever they’re deployed. #OFOproud🇺🇸

@RTSG_News How many warehouses are one disgruntled worker away from total inventory loss? Most facilities invest heavily in external security but almost nothing in the retention and wellbeing signals that prevent threats from inside.

🚨🇺🇸 BREAKING: A worker in Ontario, California sets his company’s warehouse on fire and posts message for CEO. "There goes your inventory. All you had to do was pay us enough to fucking live." Follow: @RTSG_News

Everyone's building AI agents. Nobody's building the leaders who orchestrate them. The CSO of 2027 won't just manage supply chains. They'll orchestrate fleets of AI agents alongside human teams.

@GXOLogistics Luxury is where this gets tested first because one missed condition event kills brand perception immediately. The shift from automating tasks to orchestrating decisions only works when the sensor data is live, not lagging half a day behind the shipment.

AI is changing the game in luxury supply chains, not just with automation, but through clever orchestration that makes everything work together smoothly. Luxury brands can’t afford to drop their quality standards or let the customer experience slip. Still, they have to deliver the availability and consistency today’s shoppers expect, and that’s no easy feat! Together with Vogue Business, our team unpacked what this means in practice: 🔹How logistics leaders are using AI to balance precision, scale, and sustainability 🔹Where many luxury supply chains are falling short on AI readiness 🔹How AI orchestration can protect availability across omnichannel luxury networks GXO’s Chief Technology Officer, Nizar Trigui, shared his insights in the report, which shines a light on how #AI is becoming increasingly important for #luxurylogistics. gxo.com/news_article/h…

@DCVelocity Jet fuel doubled since February and MESA capacity is still 33% below last year. The rate chart is the symptom. The routing assumptions underneath it are the exposure!

With petroleum markets disrupted by the Iran War, global airfreight pricing continued to rise in the last week of March, indicating that the price of jet fuel is replacing capacity constraints as the driver of upward rate momentum. dcvelocity.com/transportation…

George nailed the honest version of this. The show floor says resource-agnostic, the deployment floor says "which provider dropped coverage at the port gate again." Cargo IoT lives in that gap every day. Your sensor doesn't care if it's on cellular or satellite, but the operations team cares a lot when connectivity dies mid-ocean and the data goes stale for 11 days. Resource-agnostic is the right destination. We're just not close yet.

Is IoT finally becoming resource-agnostic? At #MWC26, the answer wasn’t so simple. AI, eSIM & automation are acceleratine, but complexity isn’t gone. 👉 Read the full On Trend edition: bit.ly/3Ojfpxz #IoT #MWC26 #eSIM @MWCHub @SkyloTech @LucVidalMadjar @BICS_SA @AerisM2M

The "invisible core collector" framing is the right one. Alex's stat stuck with me too: only about 50% of carriers can be auto-onboarded. That other 50% is where network quality actually gets decided. TMS as the brain, visibility as the eyes and ears. I'd add the network still needs a nose. Something that smells a problem before the shipment moves.

The Power of the Network: Enhancing Logistics Through Strategic Carrier Connection - go.shr.lc/4cwsMDP

Cargo theft losses hit $725 million in 2025. Up 60% in a single year. Average value per theft: $273,990. But the real shift isn't in the numbers. A quarter of fleet professionals now say their biggest threat isn't physical. It's strategic. Fictitious pickups. Falsified paperwork. Identity fraud at the carrier level. The threat evolved from smash-and-grab to supply chain infiltration. You don't solve infiltration with locks and cameras. You solve it with intelligence: identity verification, behavioral pattern detection, and handoff integrity at every node. carriermanagement.com/news/2026/01/2…

The IoT patching gap is where this hits supply chain hardest. We have sensors on cargo that can't be updated, legacy warehouse controllers with no security team, and now AI-generated microservices layered on top. Your Unknown No. 3 is our daily reality. Self-healing networks sound great until half your fleet runs firmware from 2019.

Cybersecurity in the Age of Instant Software schneier.com/blog/archives/…

The $20K unit cost target is the visible metric. The harder problem is the intelligence layer underneath. A Walker-series robot moving automotive parts across global chains needs real-time lane data: right part, right condition, right handoff. Getting the robot cheap is an engineering problem. Getting the data clean enough to trust it is an operations problem.

Humanoids are heading to the assembly line at Honda. 🤖 @UBTECHRobotics and Honda Trading have signed a strategic agreement to pilot Walker-series robots and autonomous logistics across global automotive supply chains. After a breakout 2025 where full-size humanoids became the company's largest revenue driver, UBTECH is doubling down on "physical AI" for the factory floor. With a long-term goal of driving unit costs below $20,000, the partnership marks a major step in moving humanoid labor from pilot programs to mass-scale industrial utility.

The speed point is what most people miss. While one company waits for tariff clarity, a competitor is already locking in a second supplier and securing terms. By the time conditions are clear, the advantage is gone. Procurement built for control in a stable environment becomes the bottleneck in a volatile one.

Margins are light, decisions take too long and competitors move first – but nothing looks broken. A 2015 procurement strategy is usually why. Here’s what’s happening. 👉 linkedin.com/posts/jimtompk… #Procurement #SupplyChain

The Gartner stat in here is the one that matters: GenAI saves individual SC workers 4 hours/week, but at the team level it shrinks to 1.5 hours with no improvement in output quality. The gains evaporate at scale. That's not a technology problem. That's a workflow architecture problem.

AI in supply chains is shifting from pilot programs to measurable P&L impact, with success determined by execution, data readiness, and operational integration rather than technology alone. scmr.com/article/ai-in-…