Engagement lead: Alvis Lazarus, Hesol Consulting   |   Sector: Automated intralogistics and third-party logistics   |   Scope: Conceptual warehouse solution design and analytical modelling

Our client is a US-based robotics-first logistics venture building a dark warehouse: a high-throughput distribution operation designed from the ground up around robotics, rather than automation retrofitted onto a manual facility. The operating concept covered pallet and case handling, replenishment distribution centre workflows, mixed case picking and cross-dock, running 24×7 under a single-tenant model.

Before committing capital, signing a lease or opening investor conversations, the leadership team needed to know what the operation would actually cost, how large the building had to be, how many people it would take to run, and where it would break.

Dark Warehouse Transformation Business Case

Automated intralogistics has moved from pilot projects to the default architecture for new distribution capacity. For an operator building from a blank sheet, getting the operating design right before capital is committed translates directly to:

1. Capital discipline through robot and equipment counts derived from modelled throughput rather than from vendor recommendations.
2. Operational efficiency from an automation mix matched to the actual flow profile, process step by process step.
3. Real estate accuracy, because facility size is the largest and least reversible commitment in the plan.
4. Investment readiness, with a cost stack and unit economics an investor can interrogate line by line.

The gap was not ambition, it was evidence. The client had a conceptual workflow and a clear view of the operation it wanted. What it did not have was a structured, traceable model connecting that concept to capital cost, operating cost, staffing, throughput, space and cost per unit handled. Hesol Consulting was engaged to build that model from scratch, and to make independent recommendations where the design required judgement rather than documentation.

Opportunity

The facility did not exist yet. That is the defining constraint of a greenfield automation business case, and it rules out the usual approach of baselining a current operation and projecting an improvement on it.

• No operating history to benchmark against, so no internal cost, productivity or throughput baselines could be drawn on.
• Robotics vendor claims were not comparable across suppliers. Productivity, maturity, cost basis and space impact are each quoted on different terms.
• Capital cost, operating cost, staffing, throughput and facility size were being estimated in isolation, so changing one assumption did not move the others and no scenario was internally consistent.
• Automation scope was still genuinely open: which process steps to automate, which to leave manual, and the volume at which that answer flips.
• Lease and fundraising decisions sat downstream of the model, which meant a spreadsheet nobody could audit would not survive diligence.

Reframed, this was not a costing exercise. It was an opportunity to build the operating logic of the business once, properly, in a form the client could own, interrogate and keep iterating on internally long after the engagement closed.

Solution

1. Operating Logic and Assumptions Architecture

• Structured the client’s conceptual inputs into a documented, version-controlled assumptions register with an explicit decision log.
• Developed the end-to-end process flow: inbound unload, dock staging, putaway, bulk and rack storage, replenishment, goods-to-person case pick, pallet build and outbound.
• Defined demand profiling, flow mix (pallet to pallet, pallet to case, case to case and cross-dock), SKU velocity segmentation and inventory norms.
• Built the assumptions hierarchy so that every downstream number traces back to a named, sourced input through a visible identifier.
• Separated user inputs, auto-calculated fields and key outputs using one consistent visual convention across every tab.

2.  Automation and Robotics Architecture

• Built an equipment master library covering autonomous forklifts, pallet and case mobile robots, pallet and tote automated storage and retrieval systems, goods-to-person stations, robotic picking and palletising arms, conveyor, dock automation, vision, dimensioning, print and apply, and inventory drones.
• Captured every technology on a common basis: throughput reference, maturity, normalised cost, handling type and space impact, which made the options comparable for the first time.
• Encoded selection logic and, just as importantly, the avoid conditions for each technology at each process step.
• Mapped seven operation-type and storage-architecture archetypes against twelve process blocks, turning technology choice into a structured decision rather than a preference.
• Derived robot and equipment quantity ranges from modelled throughput and utilisation, not from supplier guidance.

3.  Cost, Capacity and Space Engine

• Built a capital cost sub-model spanning robotics, fixed automation, storage systems, power and charging infrastructure, racking, IT, warehouse and automation control software, integration, commissioning, spares and contingency.
• Built an operating cost sub-model layering direct labour, indirect and supervisory labour, utilities, maintenance contracts, consumables, insurance, compliance and facility cost, split between fixed and variable.
• Developed a staffing model by role, process and shift, covering operators, leads, supervisors, exception handling and robot maintenance technicians, with both launch-state and steady-state views.
• Modelled throughput, capacity envelopes and utilisation to surface bottlenecks and constraint propagation, rather than reporting averages.
• Built a facility size engine deriving storage, staging, dock, charging, quality and circulation area from flow and storage requirements.

4.  Decision Layer: Unit Economics, Sensitivity and Ramp

• Created a unit economics layer producing cost per pallet handled, per pallet stored, per case picked and per truck loaded or unloaded.
• Built sensitivity views identifying which assumptions most strongly move capital cost, operating cost, throughput, staffing, facility size and unit cost.
• Modelled the ramp over the first six to twelve months: volume growth, staffing gradient, phased capital deployment and the evolution of unit economics as utilisation stabilises.
• Made the model interactive for what-if work across a base case, a stress case and a scaled high-volume case.
• Closed out with an open questions and data gaps register, a version log and a findings readout, so the handover included the known unknowns alongside the answers.

Results

Values below are withheld under client confidentiality. The capability described in each point was delivered. Only the numbers have been removed.

• A single connected operating model. Thirteen linked tabs, from navigation through to version log, with every derived figure carrying a cross-reference back to its source assumption. Built from scratch and handed over for the client to run and extend internally.
• A defensible facility size answer, derived from storage requirement and material flow rather than a rule of thumb, and split explicitly between operational area and circulation area. [WITHHELD]
• A full annual operating cost stack, decomposed into facility, maintenance, utilities, people and other categories, so the largest levers were visible rather than buried in a total. [WITHHELD]
• A storage requirement in pallet positions, derived from throughput, inventory days and peak buffer logic, which in turn drove storage system sizing and the building footprint. [WITHHELD]
• Unit economics per handling event, including cost per inbound pallet, cost per pallet stored and cost per outbound pallet, giving the client a commercial rate basis rather than only a cost total. [WITHHELD]
• A reusable robotics architecture, with selection logic and constraints documented per process step, so a change in scope could be re-evaluated without rebuilding the analysis.
• A governance trail built in: assumptions register, open questions and data gaps log, and version log, so the model stayed auditable once the engagement closed.

The engagement was structured to put a first-pass working model in the client’s hands by the end of week two, and to leave behind something more durable than a number: a transparent, traceable operating model the team could keep pressure-testing as the concept moved toward capital commitment.

A Note on What Is Not Here

This story carries no percentage improvement and no before-and-after comparison, because there was no operating baseline to improve on. The value delivered was decision quality: a defensible answer to how much, how big, how many and where it breaks, produced before the capital was spent rather than after.

To discuss a warehouse solution design, automation business case or operating model engagement, get in touch with our team. Use the contact form below: