The integrator you want has said no before

There is one question that sorts integrators faster than any credential, and almost nobody asks it.

Ask them to tell you about a job they turned down, and what made them turn it down.

An integrator who has never walked away from a project either has not been doing this long or takes everything that comes through the door. Neither is what you want holding the schedule on your first cell. The firms worth hiring have a clear picture of what they are good at and what they are not, and they can describe the edge of it without getting defensive. What you are listening for is a specific story with a technical reason in it, something like a part that could not be presented consistently, a cycle time the customer needed that the physics would not support, a finish requirement that force control was not going to hit. A firm that answers with a story like that is a firm that will tell you the truth in month four when something is not working.

The rest of the evaluation is more conventional but still worth doing properly. Ask for references from projects that resemble yours in industry, in part size and in company size, and then actually call them. A reference from a nine hundred person automotive supplier tells you almost nothing about how a firm will treat a forty person job shop, because those are different businesses being run by the same company. Ask the reference what went wrong, since something always did, and listen to how the integrator handled it rather than whether it happened.

Then go see a cell running. This is the single most useful hour you will spend, and it is remarkable how many buyers skip it. Standing in front of a working installation tells you things a proposal cannot: whether the guarding looks considered or improvised, whether the operator seems confident or is working around something, whether the changeover is as quick as the specification claimed. Ask the operator what they would change. They will tell you, because nobody ever asks them.

Find out who actually does the work. Some firms carry engineering, mechanical build, controls and installation in house. Others subcontract the controls, or the panel build, or the mechanical fabrication. Subcontracting is not a problem in itself and many excellent integrators do it, but you should know, because it determines who is standing in your building when something needs to be fixed and how long the phone chain is when it does.

Ask about service radius and response time, and ask it as a concrete question rather than a general one. Not do you offer support, but how far away is your nearest service technician to my address, and what has your actual response time been this year. A brilliant integrator six hours away is a worse choice than a good one ninety minutes away, on the day your cell is down and the order ships Friday. Ask about remote support too, meaning whether they can dial into the cell, and then ask your own IT people whether they will allow it, because that conversation is better had before installation than during an outage.

Certifications are worth something, and worth less than people assume. The A3 Certified Robot Integrator program and the manufacturer partner programs, meaning the certified integrator designations run by companies like Universal Robots, FANUC, ABB, KUKA and Yaskawa, all involve a firm meeting published criteria and being audited or trained against them. That is a real signal, and it is the reason our own bench is built from those directories. But a certification tells you a firm cleared a bar, not that it is the right firm for your job. Programs also vary in what they actually assess, so ask what the certification they hold required them to demonstrate. The good ones will know the answer in detail.

Finally, consider durability, which is the least exciting question and the one that bites hardest. This cell should run for a decade or more. Ask how long the firm has been in business, how many people it employs, and whether it will be able to support this installation in eight years. You are not looking for a giant. Plenty of ten person integrators are outstanding, and are frequently better suited to a small shop than a large firm for whom your project is a rounding error. You are looking for stability, and for the sense that the knowledge of your cell lives in an organization rather than in one person's head.