Warehouse Automation Roadmap: 12-Month Plan From Data to Go-Live

10 min read ยท 2026-10-25

A warehouse automation roadmap is the plan that takes a site from manual work to automated equipment in a safe order: understand your order and SKU data, fix the processes and systems first, build the business case, choose the right technology and vendor, integrate and test it, then go live in your quietest season and ramp up. Most mid-size projects take around a year from first analysis to a stable system, and longer for large fixed systems such as an automated storage and retrieval system (AS/RS).

The template below runs for 12 months in seven phases, each with a milestone you can check. It is written for operations directors, distribution center managers, supply chain leads, engineering managers and the IT lead who owns the warehouse management system (WMS). It also covers what vendor brochures leave out: data you need before you talk to vendors, the integration work that causes most delays, safety and workforce planning, and how to keep shipping orders while the change happens.

The roadmap at a glance

Goal: An automation system that is live, integrated with your WMS, accepted against agreed performance targets and running at planned throughput before the next peak season. Duration: 12 months

  1. Baseline and data (Weeks 1-6)

    Know your operation in numbers before you look at any equipment.

    • Pull 12 to 24 months of order lines, units and SKUs by day and hour, including peak days.
    • Run a velocity analysis to sort SKUs into fast, medium and slow movers, and note size, weight and handling needs.
    • Measure current labor hours, picks per hour, error rate and dock-to-stock time for each process.
    • Write a growth forecast with sales and finance for the next five years, including new channels such as direct-to-consumer.
    • List the constraints: building height, floor flatness, power, fire protection and lease terms.

    Milestone: A data pack with order profile, SKU velocity, labor baseline, growth forecast and site constraints, signed off by the operations director.

  2. Fix processes and systems (Weeks 4-12)

    Remove waste and system gaps so you do not automate a broken process.

    • Re-slot fast movers closer to packing and clean up location data in the WMS.
    • Standardize picking, packing and replenishment steps and write them down.
    • Check whether your WMS can send tasks to automation, or whether you need a warehouse execution system (WES).
    • Fix inventory accuracy with cycle counts until system and shelf match.

    Milestone: Re-slotting done, standard work published, inventory accuracy at the agreed target, and a decision on WMS and WES readiness.

  3. Options and business case (Weeks 8-16)

    Choose the type of automation that fits your data and prove it is worth it.

    • Shortlist two or three concepts that match your profile, such as autonomous mobile robots, goods-to-person storage or a conveyor and sortation system.
    • Model each concept with your real order data, including peak days, not averages.
    • Compare total cost of ownership, labor impact, flexibility, space and risk for each concept.
    • Present the business case with a recommended concept, payback logic and a phased plan to leadership.

    Milestone: Business case approved by the executive team, with one concept chosen and the budget released for vendor selection.

  4. Vendor selection and design (Weeks 14-24)

    Pick a vendor and lock a detailed design and contract.

    • Send a request for proposal with your data pack, functional requirements and performance targets.
    • Visit reference sites with a similar order profile and talk to their operations teams.
    • Agree the detailed layout, interfaces, software scope, service levels and spare parts plan.
    • Write acceptance criteria into the contract: throughput, accuracy and uptime the system must reach before sign-off.

    Milestone: Contract signed with acceptance criteria, detailed design approved and the installation schedule agreed.

  5. Site preparation and integration build (Weeks 22-38)

    Get the building and the systems ready while the equipment is built.

    • Complete civil and electrical work, network coverage and any changes to fire protection, with permits where needed.
    • Build and test the interfaces between WMS, WES and the equipment controls in a test environment.
    • Plan how the site keeps shipping during installation, such as temporary pick zones or a third-party overflow site.
    • Start the workforce plan: new roles, training, safety rules and early talks with staff.

    Milestone: Site ready for installation, interfaces passing end-to-end tests with sample orders, and the continuity plan signed off.

  6. Install, test and go live (Weeks 36-46)

    Install the system, prove it works and switch over safely.

    • Install the equipment and run site acceptance tests against the contract criteria.
    • Train operators, team leads and maintenance technicians on real equipment.
    • Write a cutover runbook with go and no-go criteria and a rollback plan to manual picking.
    • Go live in a low-volume period, starting with a subset of SKUs or orders.

    Milestone: Site acceptance tests passed, staff trained, and the system live for the first wave of SKUs with no rollback.

  7. Ramp-up and optimization (Weeks 44-52)

    Reach planned throughput and hand the system to normal operations.

    • Add SKUs and order types in waves and track performance daily.
    • Fix the top causes of stoppages and exceptions each week with the vendor.
    • Run a peak readiness test at forecast peak volume.
    • Compare results to the business case and close out the project.

    Milestone: Final acceptance signed against throughput, accuracy and uptime targets, and a post-project review shared with leadership.

Which automation fits your warehouse

The right technology depends on your order profile, not on what is new. Start with your data: how many order lines per day, how many units per line, how many SKUs, how big and heavy they are, and how much volume swings between normal days and peak.

Many sites start with flexible, smaller steps and add fixed systems later. The groups below are a starting point for your shortlist. Ask vendors to prove fit with your own data, and check the details of each option with them.

  • Pick-to-light or put-to-light and voice picking: low disruption, raises accuracy and speed in existing racking.
  • Autonomous mobile robots (AMRs): robots move goods or carts between people and zones, scale up by adding robots, and suit sites with changing volumes or leases.
  • Goods-to-person systems: storage systems bring totes or shelves to a pick station, which suits many small SKUs and high order lines.
  • Automated storage and retrieval systems (AS/RS): dense storage using building height, a long-term commitment suited to stable, high volumes.
  • Conveyors and sortation: move and sort cartons or parcels at high speed, common in parcel and retail distribution.
  • Packing automation: auto-baggers, box-on-demand machines and print-and-apply labelers at the end of the line.

Example: a growing distribution center adds goods-to-person

Picture a health and beauty distributor with one 200,000 square foot site, about 15,000 SKUs, mostly small items, and a fast-growing direct-to-consumer channel. Pickers walk most of their shift, peak season needs many temporary workers, and the lease runs for another ten years.

The team chose goods-to-person storage for small items and kept manual picking for bulky ones. That split came straight from the SKU data and avoided paying to automate items that did not need it.

  • Q1, Data and process: order and SKU analysis, re-slotting, inventory accuracy project, WMS review showing a WES was needed. Owner: operations director with an engineering manager. Milestone: data pack and WES decision signed off.
  • Q2, Business case and vendor: three concepts modeled with real peak data, goods-to-person chosen, request for proposal, two reference visits, contract with acceptance criteria. Owner: engineering manager with procurement. Milestone: contract signed.
  • Q3, Build: floor and power work, WES and WMS interfaces built and tested, temporary pick zone planned, training roles defined with HR. Owners: facilities manager, IT lead, HR business partner. Milestone: interfaces pass end-to-end tests.
  • Q4, Go live and ramp: installation, site acceptance tests, go-live in January with fast movers first, ramp to all small SKUs over eight weeks. Owner: operations director. Milestone: final acceptance before the next peak.

Who owns what

Automation projects cross operations, engineering, IT, finance, HR and safety. Name one project owner with authority over scope and dates, and give each workstream a lead.

Write a RACI for the main decisions: concept choice, vendor selection, interface scope, go and no-go, and final acceptance. Disputes during testing go much faster when everyone knows who signs.

  • Executive sponsor (COO or VP of supply chain): approves the business case, budget and go-live date.
  • Project owner (operations director or automation program manager): owns the roadmap, schedule, vendor relationship and acceptance.
  • Engineering manager: data analysis, concept design, layout and site acceptance tests.
  • IT or WMS lead: WMS and WES changes, interfaces, network and test environments.
  • Facilities and safety lead: building work, permits, fire protection, safety rules and risk assessments.
  • HR business partner: workforce plan, new roles, training and staff communication.
  • Finance partner: business case, total cost of ownership and post-project review.

How to present it to leadership

Leadership wants to know the problem, the payback and the risk. Put them on one slide: the business problem in numbers (labor hours, peak staffing, missed cut-offs, space), the chosen concept, the benefits and how they were modeled, the investment and payback logic, the timeline by quarter and the top three risks with their fallbacks.

Leave out the full vendor comparison and technical drawings, and keep them in an appendix. Show that you modeled peak days and growth, and say what happens if volume grows slower than forecast. Leaders are more likely to approve a plan that shows a phased path, such as starting with mobile robots in one zone, than one that bets everything on a single big system.

Metrics to track

Set the baseline in the first phase and track the same metrics through ramp-up. Use your WMS and the automation system's own reports, and agree on definitions with the vendor before testing starts.

Watch the first weeks after go-live closely. Uptime and exception rates show whether the system will reach its targets long before throughput does.

  • Throughput: order lines or units per hour at the system and for the whole site.
  • Labor productivity: units per labor hour, compared to the manual baseline.
  • Order accuracy: errors per thousand orders, from returns and customer claims.
  • System availability: share of planned run time the system was working, from the controls software.
  • Exceptions: totes or orders sent to manual handling, by reason.
  • Order cycle time: time from order release to ready for shipping, against carrier cut-offs.

Risks and dependencies

Most automation delays come from integration, site work and people, not from the equipment itself. Keep these risks in a risk log with an owner and a fallback, and review them in every weekly project meeting.

Also plan for the day the system stops. Write down how you will ship orders if it is down for an hour, a shift or a day.

  • Interfaces between WMS, WES and controls are late: owner IT lead. Fallback: start interface testing early with simulated equipment and freeze the interface scope.
  • Building work or permits slip: owner facilities lead. Fallback: start permits as soon as the layout is fixed and build buffer into the schedule.
  • Go-live clashes with peak: owner project owner. Fallback: move go-live to after peak rather than going live under pressure.
  • Staff resistance or turnover: owner HR business partner. Fallback: involve operators early, explain new roles and train on real equipment.
  • Vendor support gaps after go-live: owner project owner. Fallback: service levels, spare parts and on-site support written into the contract.

How to adapt it

For a smaller site, or one with a short lease, focus on flexible options such as mobile robots, pick-to-light or packing automation. They can often be piloted in one zone in a few months, then scaled. Keep the same order: data, process fixes, business case, then equipment.

For a large multi-site network, run the roadmap as a program. Build a standard design at one pilot site, prove it, then roll it out in waves. For fixed systems such as AS/RS or large sortation, allow more time for design, building work and testing, and plan the go-live a full season ahead. In regulated sectors such as pharmaceuticals or food, add validation and traceability steps, and confirm the requirements with your quality team.

Common mistakes to avoid

  • Talking to vendors before you have your own data lets them shape the problem, so finish the order profile and SKU analysis first.
  • Automating a messy process locks the waste in, so re-slot, standardize work and fix inventory accuracy before you design the system.
  • Sizing the system on average days leaves you short at peak, so model peak days and five-year growth with real order data.
  • Treating integration as a vendor task leads to late surprises, so give your WMS or IT lead ownership of the interfaces and test them early.
  • Going live just before peak season puts your busiest weeks at risk, so plan go-live in a low-volume period with a rollback plan.
  • Leaving the workforce plan until the end breeds fear and turnover, so involve operators early and define new roles and training during the build.

Frequently asked questions

How long does warehouse automation take to implement?

Smaller, flexible systems such as pick-to-light or mobile robots can be piloted in a few months. Mid-size projects often take around a year from first analysis to a stable system. Large fixed systems such as AS/RS usually take longer, so check timelines with vendors and reference sites.

What is the first step in a warehouse automation roadmap?

Collect and analyze your own data: order lines, units per line, SKU velocity, item sizes, labor hours and peak days. That data tells you which kinds of automation fit. It also gives you the baseline you need for the business case.

Do I need a new WMS before automating?

Not always. Check whether your WMS can send tasks to the equipment and handle the volumes you plan. Some sites add a warehouse execution system (WES) between the WMS and the equipment instead of replacing the WMS.

How do you build a business case for warehouse automation?

Compare the current cost and capacity of the operation with each automation option over several years, using your real order data and growth forecast. Include labor, space, accuracy, peak staffing and running costs such as maintenance and software. Show the payback logic and what happens if volumes grow slower than planned.

When is the best time to go live with warehouse automation?

Go live in your lowest-volume period, well before peak season, so the team has time to ramp up. Start with a subset of SKUs or orders and add more in waves. Keep a rollback plan to manual work until the system has proven itself.

Generate this roadmap with AI