The protocol is rarely as fixed as it looks — hidden requirements are what stretch the timeline.
Automation projects usually run late because the process changes or hidden requirements appear after the project starts. The original estimate may be reasonable, but new labware, new steps, edge cases, or biological behaviour can add weeks or months.
A project runs late when the work needed to make the process reliable turns out larger than the work described at the start. In biology this happens because protocols aren't as fixed as they seem. Automation teams aren't only building motions — they're turning a real scientific workflow into a repeatable physical process, and every unclear step, exception, or hidden assumption has to be discovered and handled.
A manual protocol can look simple until it's automated. A scientist instinctively adjusts for bubbles, clogged tips, unusual liquid behaviour, or a meniscus that looks wrong; a robot has to be given a way to detect and respond to each. Magnetic beads, for instance, may clog pipettes more often than expected — the commands are correct, but the process needs an added check to confirm volume.
One issue like that might add a week. A dozen of them can add months.
Before handing a workflow to an automation team, run it as if a human were pretending to be the robot. Record every movement, decision, loading step, and exception. The more specific the process is before engineering starts, the less room hidden requirements have to stretch the timeline. Related: how long deployment takes and why projects fail.
