Over the winter we sat down with seven procurement leads across European tier-1 manufacturers — two in automotive, two in FMCG, one in aerospace, one in pharma and one in white-goods. The goal was a simple question: how do you actually decide which robotics vendor goes from pilot to production line?
The answers were remarkably consistent, and in some cases surprising. The technical bar is lower than most founders assume. The commercial, compliance and service bars are substantially higher.
The pilot is a procurement filter, not a technical test
Every single procurement lead we spoke to described the pilot phase as primarily a way to evaluate the vendor's operational discipline, not the robot's capability. Can the vendor show up on the agreed date? Do they file change orders correctly? Does their safety documentation survive a regulatory audit without rework?
We have never rejected a supplier because the robot was not good enough. We have rejected plenty because the quarterly business review was late.
What actually gets measured in a pilot
- OEE (overall equipment effectiveness) versus the line it is replacing or augmenting
- Mean time between interventions, with interventions defined by the customer, not the vendor
- Safety-incident count and near-miss count across the pilot window
- Integration effort measured in engineering-hours the customer had to contribute
- Supplier responsiveness measured in hours from ticket open to first substantive response
The thing to note is that only the first two of those are about the robot. The rest are about the company building the robot.
Pricing discipline is the fastest way to lose the deal
Two of the seven procurement leads we interviewed independently described cases where a vendor re-priced between pilot and production — typically because the pilot had been sold at a loss and the vendor needed to recover margin at scale. In both cases the vendor was removed from the supplier list and the procurement team went back to the runner-up.
If you are a founder, the implication is uncomfortable: the pilot price has to be within 15% of the production price, and both have to be profitable. The tempting 'loss-leader pilot' is a strategy that works for SaaS and almost never for industrial hardware.
The compliance stack is non-negotiable
Across the seven interviews we heard the same short list of requirements: CE and Machinery Directive compliance documentation, ISO 10218 and ISO/TS 15066 for collaborative applications, a TÜV or equivalent safety assessment, and full GDPR documentation for any data collected from the line.
For pharma and food, add to that a documented cleaning and validation procedure that a compliance auditor can sign off on. For aerospace, add AS9100 and a traceable change-control process. Founders who show up to pilot meetings without this stack ready lose six to nine months while they assemble it.
The role of the integrator
In six of the seven manufacturers we talked to, the robot vendor does not contract directly with the factory. A system integrator — usually one of the incumbents like KUKA Systems, Dürr or ABB Robotics Integration — owns the customer relationship. The founder's first commercial conversation is therefore not with the factory but with the integrator.
Timelines are long but predictable
The median timeline we saw from first technical conversation to a production-line purchase order was 17 months. Once the vendor was approved onto the supplier list, subsequent orders landed in under four months. That gap — 17 months versus 4 months — is the real reason repeat customers are so much more valuable in industrial robotics than in almost any other category.
For investors, the right diligence question is not 'how big is the pipeline' but 'how many accounts have crossed month 17.' That is where the margin really compounds.