MODEX (and PROMAT) have always felt like a World’s Fair for supply chain—a place where you can watch the “future” perform live: robots picking, conveyors routing, shuttles flying, arms assembling, and demos running flawlessly on repeat. For years, the headline story was hardware: faster, more autonomous, more mechanized.
This year, the story felt different.
Below are three trends that stood out to me at MODEX 2026—each with the potential for true impacts on distribution center performance, investment decisions, and how we should evaluate “innovation” going forward.
1) Hardware automation cools down, but software automation is heating up
Historically, the show floor has been dominated by physical automation: robots, AS/RS’s, sortation, goods-to-person systems, and all the mechanized material handling. It’s still impressive—and in the right use cases, it’s transformative.
But the “robotics frenzy” feels like it has matured.
Not because automation doesn’t work—In plenty of scenarios it does. But many operators are now living through the less-advertised reality:
- Throughput isn’t meeting expectations (sometimes materially below plan)
- “Modular” systems are harder to scale than promised
- Peak volumes reveal brittleness
For many clients, the work today isn’t buying automation—it’s stabilizing it. Especially during peak when systems don’t behave the way the business believed they would. (You already know what I mean if you saw Steve Mulaik’s MODEX 2024 talk, “Common Pitfalls in Automation Projects: How to Avoid Your Project Turning into a Dumpster Fire.”
Meanwhile, software is everywhere—often pitched with the same promises that hardware used to lead with:
Less labor, higher throughput, faster onboarding, “plug-and-play,” and now—SaaS economics.
The buzzwords can be dizzying: bolt-on, end-to-end, AI-powered, autonomous. History doesn’t repeat, but it certainly rhymes.
2) The Word of the Day: Orchestration
At MODEX 2022, I heard the term “orchestration” zero times.
At MODEX 2024, I heard it occasionally, but—mostly in conversation.
At MODEX 2026, it’s printed everywhere: posters, brochures, product pages, session titles.
And I suspect that by MODEX 2028 (or 2030), it will be unavoidable.
Why?
Because the industry is bumping into a ceiling: you can buy more machines, but you can’t brute-force flow. When volume spikes or variability increases, operations don’t fail because they lack equipment—they fail because the system can’t coordinate decisions fast enough across constraints.
This is where the concept of orchestration becomes more than a buzzword.
Orchestration is “The Goal” for distribution centers
If you’ve studied industrial engineering, Lean, Six Sigma, or process optimization you’ve probably encountered The Goal by Eliyahu M. Goldratt. It’s written like a novel but teaches a foundational principle: constraints (bottlenecks) govern throughput.
Manufacturing embraced these principles decades ago, distribution and fulfillment are catching up.
In DC terms, orchestration is the practical application of constraint-based thinking to:
- release work at the right time (not just as fast as possible),
- balance labor dynamically,
- manage exceptions proactively,
- keep flow moving through the true bottlenecks.
From “waves” to flow: why waveless matters
A great example is the shift from traditional waving (batching) to waveless decisioning.
In manufacturing, moving from batch-and-queue to single-piece flow changed everything. In fulfillment, “waveless” is a similar evolution: rather than grouping orders into waves because the system needs batches, modern orchestration can make release decisions continuously based on constraints, priorities, and real-time conditions.
Some providers have understood this for a long time, but it’s historically been tough to market because it’s not as visually flashy as a robot. Others have implemented pieces of it, but not the full operating model. And some legacy architectures struggle to support real-time decisioning without major redesign.
At MODEX this year, the most compelling conversations I had were with people who talk about orchestration: not as a feature, but as a control philosophy.
A note on “bolt-on” orchestration
The strongest orchestration approaches I saw weren’t necessarily “rip and replace.” They were lightweight, configurable layers designed to integrate with what’s already there—while improving release logic, labor balancing, exception handling, and visibility.
The pre-eminent example of this is Fulfillment Engines. Check them out here to learn more!
3) “AI” is Wildly Overused — and Often Misunderstood
We have to talk about AI—because the term is used very loosely and incorrectly in the marketing of AI products today.
Back in the 2000s, we called the computer opponents in video games “AI,” even though they were basically pre-programmed behaviors: if X happens, do Y. That’s not AI, it’s an algorithm—rules and logic applied to known inputs.
Generative AI and modern machine learning are different. They’re probabilistic. They can generalize, create, and respond to novelty—but they can also be:
- less reliable (not guaranteed to produce the optimal result every time),
- harder to explain (“black box” behavior),
- heavier computationally, which can matter for real-time decisions.
Here’s the issue: many software companies market “AI” when the product is primarily algorithmic decision logic—because “AI” sounds cutting-edge.
And in many operational contexts, algorithmic optimization is exactly what you want. If the goal is to make repeatable, provably-good decisions under constraints, deterministic math can be more reliable than probabilistic output— especially time-sensitive orchestration.
So when you hear “AI,” the practical question isn’t “Is it AI?”
It’s:
- What problem is it solving?
- How does it make decisions?
- How does it handle exceptions and edge cases?
- What does success look like at peak?
- Can it prove improvements in throughput, labor, and service?
In Closing
MODEX 2026 felt like a shift from “look what the robots can do” to “how do we control the system?”
Hardware still matters. But the competitive edge is increasingly in decisioning, orchestration, and operational control—especially under variability and peak pressure.
If you’re evaluating software automation, hardware automation, or broader operations design challenges, this is exactly the space where we spend our time at Argon & Co.—helping teams separate demos from deployable value.
Connect with Zach Hess on LinkedIn
Share:
Zach Hess
Three Industry Trends at MODEX 2026
