The warehouse-automation conversation is dominated by greenfield showcases: purpose-built sheds with 40-foot ceilings and seven-figure robot fleets. But roughly three-quarters of warehouses run with no automation at all, and nearly all of them are existing buildings with existing problems. For that majority, the real question is not which flagship system to buy — it is whether retrofit automation pencils at all.

Why brownfield is a different problem

A retrofit inherits constraints money can't fully fix: low clear heights, column grids that fight racking layouts, floors that won't take point loads, docks in the wrong places, and a live operation that can't stop for construction. Every one of these discounts either the technology choice or the throughput it can reach.

The retrofit technology ladder

  1. First rung — software and process. Slotting optimization, labor management, voice picking. Capex under $100K, paybacks often under a year. Skipping this rung to buy robots is the most common brownfield mistake.
  2. Second rung — flexible robotics. AMR fleets (goods-to-person or transport) deploy into existing racking in 6-12 weeks, scale by the unit, and tolerate building quirks. Typical entry $500K-$2M or RaaS monthly.
  3. Third rung — modular fixed automation. Shuttle or cube systems (AutoStore) can retrofit into surprisingly tight boxes when density is the binding constraint — if the floor and columns cooperate. $2M-$8M with 2-4 year paybacks.
  4. Fourth rung — full AS/RS. Rarely the brownfield answer. When a retrofit evaluation lands here, the honest comparison is usually against a different building.

The brownfield traps

Automating a broken process (the robot makes the chaos faster), underestimating integration with a 15-year-old WMS, skipping the floor survey (point loads are a real constraint), and modeling payback on the demo-hall throughput instead of a congested Tuesday in your building.

The operator's list

Sequence matters more than selection: rung one before rung two, pilot one zone before the building, and require the integrator to model YOUR building's constraints in writing — floor, columns, WMS vintage, and live-ops cutover plan — before any equipment quote.

Independent automation industry analysis — not investment or procurement advice. Capex decisions warrant site-specific engineering diligence.