Reported outcome
22,500 picks per day in one shiftKPI Solutions reports that Outdoor Network's complete multi-vendor automated facility can handle up to 22,500 picks per day, process orders in one shift and reduce labour dependence by nearly 50% compared with the previous system. The result covers the whole facility and cannot be attributed to the ten Geek+ AMRs alone.
Case context
Why this case matters: a distributor does not prove an after-sales or replacement-parts operation by buying one robot fleet. It has to connect storage, oversized-item movement, software, packing, exception handling and local support into one serviceable system. Outdoor Network, a US distributor of vehicles, OEM parts and accessories, worked with KPI Solutions on a multi-vendor automation project at its Albany, Georgia distribution centre. Ten Geek+ autonomous mobile robots handled one defined part of that design: moving carts of oversized inventory between receiving, storage and order consolidation.
Case boundary: this is a partner-specific analysis of Geek+’s role inside KPI Solutions’ wider system for Outdoor Network. KPI reports that the complete operation can handle 22,500 picks per day, processes orders in one shift and reduced labour dependence by nearly 50% compared with the previous system. Those are system-level claims covering AutoStore, Geek+ AMRs, conveyors, packing equipment, software and a larger facility. They are not a controlled measurement of the ten AMRs alone, and they do not disclose project profitability or the service-level performance of every component.
The parts-distribution problem came before the robot choice
KPI Solutions’ Outdoor Network case study describes inventory split between two facilities. Small parts used ageing goods-to-person carousels, while larger items were stored separately. More than half of orders were single-line, the warehouse-management system was outdated and unsupported, staffing two shifts was difficult, and parts shortages increased the need for safety stock. The operating problem was therefore not simply walking distance. It combined storage capacity, mixed item sizes, order consolidation, software continuity and labour availability.
Outdoor Network asked for more peak capacity, additional safety-stock storage, one-shift operation and a move into a larger refurbished facility. That requirement matters to an importer or distributor because it gives every technology a bounded job. A mobile robot should not receive credit for dense small-parts storage, packing or warehouse controls that were supplied elsewhere.
Evidence-supported implementation sequence
- Model the real order and inventory profile: KPI says its team analysed operating data, surveyed the existing facilities and compared layouts, throughput, peak volume, scalability and lead time.
- Separate small parts from oversized flows: the selected design used a dense AutoStore system for small parts and static shelving or pallet rack for larger items.
- Assign AMRs one visible transport role: ten Geek+ AMRs collected 75 wheeled carts from staging, moved oversized inventory to putaway areas and brought oversized picks to consolidation.
- Integrate the surrounding equipment: conveyors, right-sized packing equipment, autobaggers and cartoning systems moved and packed other order types.
- Keep control logic with an identified owner: Outdoor Network’s internal team developed its WMS and automation controls, while KPI provided interface training, simulation tools and design support.
- Measure the complete production operation: the published result covers picks, shifts, labour dependence, receiving rate and order-processing time across the integrated facility.
What the reported result can and cannot prove
KPI reports capacity of up to 22,500 picks per day, single-line picking at an average 400 per hour, multi-line picking at 300 per hour and receiving at nearly 200 items per hour. It also reports that most in-stock orders can be fulfilled, packed and loaded onto a trailer within 15 minutes of receipt. An AutoStore account repeats the 22,500-pick, one-shift and roughly half-labour-dependence results and includes comments from Outdoor Network managers that the installed system surpassed simulation targets.
The evidence is useful because it describes a live parts-fulfilment operation and identifies where Geek+ AMRs sit in the flow. It does not isolate how much throughput, labour change or processing time came from Geek+, AutoStore, the new building, the WMS, other equipment or changed work design. Published sources also use different wording around the percentage comparison with the old peak, so this case keeps the concrete 22,500-pick figure and does not convert the percentage into a stronger causal claim.
Responsibility map for a serviceable system
| Operating area | Visible responsibility in the sources | What a distributor still needs to confirm |
|---|---|---|
| Site and system design | KPI analysed data, compared layouts and integrated the multi-vendor automation design. | Performance acceptance, change control, subcontractor accountability and responsibility when interfaces fail. |
| Oversized-item transport | Ten Geek+ AMRs moved wheeled carts for putaway and delivered oversized picks to consolidation. | Fleet availability, charging design, recovery from blocked routes, software support, spare robots, batteries and parts. |
| Dense small-parts storage | AutoStore equipment handled more than 80,000 stored SKUs through bins, robots and goods-to-person ports. | Capacity assumptions, redundancy, preventive maintenance, component stock and expansion limits. |
| Software and interfaces | Outdoor Network’s internal team developed the WMS and automation controls; KPI supplied training, simulation and design input. | Interface ownership, version compatibility, cybersecurity, backups, incident escalation and recovery-time targets. |
| Parts fulfilment and service continuity | The integrated operation combined stocked and oversized items and consolidated orders for outbound shipment. | Fill rate, backorder ageing, wrong-part returns, warranty claims, emergency orders and support coverage outside ordinary hours. |
A parts-and-lifecycle gate before scale
The transferable lesson is not to copy the equipment list. It is to make each supplier’s role measurable inside a complete parts-service flow. A distributor evaluating warehouse robotics should reproduce one representative operation before extending the system to another facility or committing to a larger fleet.
- Freeze the SKU population, item dimensions, order mix, peak profile, backorder rules and oversized-item exceptions used in the design model.
- Define separate acceptance measures for storage, AMR transport, consolidation, packing, software interfaces and end-to-end order completion.
- Run witnessed tests for blocked aisles, unavailable stations, low battery, network loss, unreadable labels, missing inventory and mixed small-plus-oversized orders.
- Identify who owns the WMS, robot-management software, interfaces and production changes, including approval and rollback authority.
- Agree local stocks of batteries, wheels, sensors, chargers and other critical spares, plus repair turnaround, remote access and escalation hours.
- Measure fill rate, pick accuracy, queue time, manual interventions, downtime, recovery time, labour hours and returns against the frozen baseline.
- Scale only when the system-level result and each critical component’s service plan pass; otherwise revise the design or stop without disrupting the existing parts operation.
ChinaBrandPath reading
Outdoor Network’s case shows why a B2B technology partnership needs a bill of responsibility, not a single vendor headline. The Geek+ AMRs had a narrow, observable job inside a broader OEM-parts fulfilment system, while KPI integrated the project and Outdoor Network retained control of core software. Reported production results make the operating sequence worth studying, but they do not prove that the same fleet, facility or economics will transfer to another distributor. Before scale, the relevant decision is whether the complete workflow and its lifecycle-support obligations have been tested together under real failure and recovery conditions.
Cooperation process
KPI analysed Outdoor Network's order, inventory and facility data; the parties separated dense small-parts storage from oversized-item flows; ten Geek+ AMRs moved 75 carts between receiving, static storage and consolidation; other vendors supplied storage, conveyors and packing equipment; Outdoor Network built the WMS and automation controls with KPI support; the integrated production operation was then measured.
Responsibilities of both sides
Geek+ visibly supplied AMRs for oversized-item cart movement. KPI visibly led analysis, system design, integration, simulation and interface training. Outdoor Network operated the parts business and developed its WMS and automation controls. Public sources do not fully allocate acceptance liability, software version control, cybersecurity, preventive maintenance, critical spares, repair turnaround, remote access, warranty, incident escalation or disaster-recovery duties.
Result limitations
The reported throughput, shift and labour results cover a larger building and a multi-vendor system that includes AutoStore, Geek+ AMRs, conveyors, packing equipment, software and redesigned work. The sources do not isolate the AMRs' contribution, disclose project cost or profitability, or provide component-level uptime and service data. Percentage descriptions of improvement also use differing baselines, so the case retains the concrete reported operating figures.
What an importer can use to decide
Before scaling warehouse robotics, reproduce one representative parts-fulfilment flow and assign measurable acceptance, interface, spare-parts, repair, recovery and escalation duties to each party. Advance only when the end-to-end operation and every critical component can meet the frozen throughput, accuracy, downtime and recovery criteria without disrupting the existing service obligation.
Source, method and review
Primary source: KPI Solutions Outdoor Network case study; Geek+ implementation account; AutoStore customer account; KPI/MHI case PDF
Source checked:
Last editorial review:
Editorial owner: ChinaBrandPath editorial team
Reported figures remain claims of the named source unless an independent source is explicitly cited. The analysis separates those claims from ChinaBrandPath’s operational interpretation.