Managing the operational layer where autonomous instrumentation, AI-driven analysis pipelines, and human researchers intersect. Procurement of advanced microscopy and robotics systems. Integration of machine-learning workflows into experimental design. The quiet redesign of what a lab technician's career looks like when half the bench work is done by models.
The equipment is the easy part
Procurement is logistically demanding but conceptually solvable. The hard part is what happens after the instrument lands on the floor: who runs it, how it is shared, how data flows from it into pipelines that nobody on the original purchase order anticipated, how the people who used to do the manual version of this work find a new role that is not a demotion.
Redesigning the bench
An autonomous lab is not a human lab with a robot bolted on. The roles change. The skill mix changes. The decision rights change. Most labs underestimate how much organisational design work this requires, then end up with expensive equipment underused because no one redesigned the workflow around it.
Most of my time goes into the human and organisational adaptation. The equipment vendors handle the equipment.