I'm going to say something that won't make me popular with the sales side of this industry: most downloadable warehouse control system specification guides are obsolete the day they're published, and if you build your vendor shortlist from one in 2025, you are filtering out the wrong companies.
I say that as someone whose job is procurement, not engineering. I coordinate the RFQ. I read the specs. I'm the person who has to point out that the "integration" a vendor described in their brochure contradicts the interface section in their own proposal. For the last few years I've been on the admin-and-purchasing side of warehouse automation projects — a system logistics AGV rollout at one company, a bulk warehouse execution system refresh at another, and one warehouse control system wholesale pilot I'd rather not talk about (more below).
Here's what I've learned after sitting on both the buyer side and the "please brief the vendor" side of these decisions.
The WCS / WES / fleet-manager line has moved, and your spec template hasn't
Five years ago, a warehouse control system specification guide drew a clean stack: WMS on top, WCS in the middle, PLCs at the bottom, and a nice little box for "AGV interface" off to the side. That was accurate circa 2020. It is not accurate now.
What changed: AGV and AMR fleets went from a side feature of the WCS to the center of it. A modern system logistics AGV deployment — whether it's 20 robots or 200 — needs a control layer that does traffic management, order batching, and dynamic reassignment. Increasingly that logic lives in a WES, not a WCS. So when someone hands me a spec that says "WCS shall manage the AGV fleet per section 4.3," I want to know which 4.3 they mean (i.e., which subsystem actually does the managing), because two vendors on my last shortlist put those exact words into completely different systems. One meant "the WES tells the WCS where to send the robot." The other meant "the WCS's own fleet module." Both are defensible. Both produce very different scope-of-work meetings.
System logistics — the discipline of making WMS, WES, WCS, and AGV fleets behave as one system — is precisely what the old template does not cover. If you're negotiating a specification in 2025, ask the vendor to show you, on one slide, which system does traffic control, which one does order release, and which one the robot's onboard software actually talks to. If they can't answer on one slide, they haven't built what the spec describes.
The "wholesale" assumption breaks the specification before you even send it
Old templates assume a single-integrator model: one vendor, one throat to choke, one set of standards. That model still works for plenty of companies. But warehouse control system wholesale procurement — buying the control layer the way you buy conveyor sections or OEM hardware — is now a real option for mid-size operations, and the template doesn't allow for it.
The template will ask you to describe the acceptance test, the warranty terms, and the escalation path. It will not ask you what happens if the WES comes from Vendor A, the AGV fleet from Vendor B, and the WCS from Vendor C. But that's exactly the structure a lot of buyers end up with, especially once OEM or private-label flexibility enters the WES conversation. I've only worked with domestic vendors on these projects, so I can't speak to how the multi-vendor idea plays out internationally — but domestically it works, and the old guide actively discourages it.
The fix is to add a single page to your own template: a responsibility matrix for the control layer. Not a RACI in the corporate sense — just a grid with every subsystem you're buying on the left, and next to it, the name of the vendor accountable for each interface. If two rows share the same name, fine. If one row is blank, that's your next meeting.
The most polished specification guide is the one I trust the least
This is the counter-intuitive one, and it's the point I most expected to be wrong about. When I built my shortlist for the 2023 pilot, I ranked vendors partly by how complete their warehouse control system specification guide was. The most polished guide won the shortlist. It had sections, sub-sections, decision trees, and a 30-page appendix. It was also, in hindsight, describing a product that hadn't meaningfully changed since 2019 (this was back in 2023, and the last update to the guide had been 2021).
A vendor whose spec hasn't been touched in three years is telling you something useful — just not what they want you to hear.
What I should have done — and what I do now — is ask each vendor for the change log of their specification document, not the document itself. Then ask: what changed in the last 18 months, and why? A vendor whose spec was updated because a customer demanded a new feature is telling you something useful. A vendor whose spec hasn't been touched since 2021 is also telling you something useful, just the opposite.
I went back and forth between the polished vendor and a smaller one for about two weeks. The polished vendor offered a single, integrated stack. The smaller one offered a WES that worked with three AGV vendors I already had on the radar. On paper, the polished stack made more sense — fewer interfaces, one invoice, one escalation list. But my gut said a best-of-breed control layer would survive renegotiation better. I chose the smaller one. The implementation ran three weeks long, and I still think it was the right call.
What I ask vendors instead
This is roughly what I now use in place of the downloaded guide. It's not comprehensive, and it's specific to the control layer — but it filters more than the guide does.
- Interface list, in writing. Every interface (WMS↔WES, WES↔WCS, WCS↔AGV fleet, WCS↔PLC), the protocol, and the document that defines it. If the vendor can't share interface documents at RFP stage, they aren't ready to bid.
- Change log of their own specification. Last 18 months, minimum. If they can't produce one, that's data too.
- A named operations reference, not a CIO reference. You want to talk to the person who lives with the fleet manager at 2 a.m., not the person who signed the deal.
- The version of their own guide they were using in 2022. Compare the two. What disappeared, and why?
That's the bar. It's lower than the downloaded guide, but it's real, and it lets you compare vendors without pretending you're an engineer.
This was accurate as of Q1 2025. Warehouse automation moves fast — verify current rates, standards, and vendor roadmaps before you budget. For reference, AGV safety requirements are typically anchored to ANSI/ITSDF B56.5, and if you're buying through an OEM arrangement, ask the OEM which version of the standard their system was validated against.
The pushback I expect
Two objections, both fair.
"We like a single-vendor stack. One contract, one escalation path." Fair. If you're going that route, the guide still doesn't help you — it just helps you run the RFP inside the assumption you've already made. Use the responsibility matrix anyway; you'll still need it if the single vendor subcontracts half the scope, which is more common than the sales deck suggests.
"You're not an engineer. Why should I listen?" Because I'm the one who signs the paperwork. The engineer tells you what the system should do. The procurement person tells you whether the vendor can prove they did it. Those are different jobs, and the specification guide was written for the second job — badly, by people who haven't actually done it.
My experience here is based on a handful of mid-size projects, mostly in North America. If you're running a 200,000-square-metre facility or an international multi-site rollout, your experience may differ — the interface list will change, but I suspect the change-log test still works.
Bottom line: the warehouse control system specification guide is a starting point for a vendor conversation, not a filter for your shortlist. Update your own template every twelve months. If you haven't changed it since 2022, neither has your understanding of the market — and the vendors on your shortlist have already noticed.