BiopharmaLive
A biopharma warehouse stopped counting bins by hand.
We gave a biopharma warehouse live bin occupancy and daily reports from its own systems, and the morning count that took about three technician-hours has gone.
A biopharmaceutical manufacturer
- About 3 hoursof technician time returned every day, now the morning bin count has gone
- Twice the volumethe warehouse can now plan for, without adding headcount
- 2 hours to 30 minutesthe target for each shift lead’s daily reporting
- €70,000–118,000a year in avoidable labour: the cost the work set out to remove
The situation
A pharmaceutical warehouse at a large-molecule manufacturing site was losing more than five hours of productive time every day to manual work. The information it needed already existed, in the warehouse management systems, in email and in people’s heads.
Three technicians counted free bin spaces in the high-bay area for about an hour every morning. Shift leads spent around two hours a day working through up to 20 reports to build a picture of the operation. Urgent stock transfer orders were tracked in email, and handover between shifts was verbal.
The site costed the avoidable labour at €70,000 to €118,000 a year. And production volume was projected to double, which that way of working could not absorb.
What we built
We modelled locations, inventory and stock transfer orders once, as a single shared model of the warehouse. Both capabilities draw on it, and it’s the foundation for whatever the warehouse builds next.
Live occupancy. Bin-level occupancy for the high-bay area, drawn from SAP Extended Warehouse Management and Oracle Warehouse Management through Databricks, and shown on floor-mounted displays.
A reporting assistant. A dashboard with generated summaries that answer the questions shift leads ask every day: which inbound and outbound deliveries are overdue, how big the picking and putaway backlog is, what is due today, tomorrow and this week, which cycle counts are open or overdue, and how many alerts are open.
The warehouse team lead and the shift leads designed it with us in co-design sessions. It went through user acceptance testing and ran beside the existing routine before it went live.
The decision that mattered
The platform reads from the warehouse systems and writes nothing back to any of them, so it informs the shift and the shift lead still decides. For this site and this use it’s classified non-GxP and decision support only; another site, or another use of the same data, would need its own assessment. Every prompt, grounding source and response is logged, and people sign in through the site’s own identity provider.
Where it stands
The morning count has gone, returning about three technician-hours a day. The reporting assistant is aimed at cutting each shift lead’s daily reporting from two hours to 30 minutes.
The warehouse now has planning headroom for the projected doubling in volume without adding headcount, and the system is covered by a contracted support agreement.