For the 90% of US warehouses that thought automation was out of reach, Lumper reframes robotics as labor you hire, not equipment you buy.
By the Lumper team
U.S. warehouses spend roughly $75 billion a year on labor and yet only about 10% of them have any automation at all. Not because the technology does not exist. Because for a decade, nearly every warehouse robot on the market was sold the same way: as a capital project.
Buy the equipment. Retrofit the facility. Sign an integrator and wait six months. If that is the only version of automation on offer, most operators are going to stay manual — even when labor is their single biggest line item, even when they cannot keep a shift staffed.
The barriers are real and well-documented. Warehouse automation projects start around $200,000 for targeted deployments and routinely exceed $500,000 for larger systems. An Automated Storage and Retrieval System can run anywhere from $70,000 to more than $1.5 million. Robotic picking systems fall between $100,000 and $1 million or more, depending on scope. Those figures do not include the facility retrofitting, the integration contract, or the six-to-eighteen months of disruption before the first case is picked.
For distribution centers and wholesale operations running on manual labor, that is not an investment decision. It is a barrier to entry that was never designed to come down.
We are launching Lumper because that framing was the problem, not these operators.
The automation gap isn't a technology problem — it's a pricing problem
Ask a warehouse operator why they have not automated and you rarely hear "robots don't work." You hear a version of the same sentence instead: "We're not ready for automation."
What they actually mean is: the $500K crane system, the six-month facility retrofit, and the integration contract were never built for them. Automation got packaged as a procurement decision made once a decade by a capital committee, when what these operators actually needed was a way to cover today's picking and loading shifts.
The technology itself is not the constraint. Autonomous Mobile Robots (AMRs) have been the subject of sustained research and development over the past decade, and the results are mature. Unlike earlier Automated Guided Vehicles, which required wires or magnets embedded in the floor to navigate, modern AMRs use SLAM mapping, LiDAR, and AI-based navigation to move through dynamic warehouse environments without any pre-installed infrastructure. The hardware is ready. The pricing model is what shut most operators out.
The pattern that emerges from operators evaluating robotics is consistent: integration complexity is a genuine concern, particularly when an existing WMS does not cleanly hand off pick lists to an unfamiliar robotic system. Seasonal volume swings compound the problem further. A facility running 5,000 orders a day in steady state but expecting 75,000 orders over four days during peak cannot afford to size a CapEx investment for the peak and carry idle machinery the rest of the year. Temp agencies fill the gap in theory, but as operators have found in practice, agencies cannot always keep up with demand, and quality control degrades when staffing is stretched.
So we built the opposite. Lumper deploys autonomous warehouse robots the way you hire temp workers — by the job, priced by the pick, with nothing upfront.
What we're shipping
Lumper puts autonomous robots on your existing floor to handle the three physically hardest jobs in a warehouse.
Case picking — mixed-SKU boxed goods up to 65 lb, picked out of the racks you already have. No retrofitting, no change to your layout, driven off the pick list your WMS already produces.
Trailer loading and unloading — the highest-turnover, highest-injury job on the dock, handled by robots with remote operators standing by for edge cases.
Palletizing — outbound mixed-SKU pallets built autonomously, plugged into your existing ERP, WMS, or WES.
The same constraints apply across all three: deploy in hours, pay per pick, scale to volume instead of headcount.
Each of these tasks shares a defining characteristic: they are physically demanding, high-repetition jobs associated with the injury rates and turnover figures that make warehouse floors difficult to staff at adequate levels. Case picking is a documented bottleneck in warehouse fulfillment, and trailer loading and unloading consistently registers as the highest-turnover role on the dock. The case for automating these roles is not theoretical — it is reflected in the cost of recruiting, retraining, and working around the gaps when those roles go unfilled.
What Lumper does differently is not the task coverage. It is the deployment model. WMS integration is the source of friction that most stalls robotics projects: operators know which system runs their floor, but compatibility between that system and a new robotic solution requires validation before any hardware arrives. Lumper's robots work off the pick list your WMS already produces, which removes the integration project as a precondition for starting.
How it works
Because there is no capital project to stand up, the deployment path is deliberately short:
Assess — we evaluate your picking, loading, and palletizing workflows and find where robotic labor pays for itself first.
Deploy in hours — the robots drop into your current layout and start working off the systems you already run.
Scale by the pick — you pay only for the work performed, and capacity follows your throughput, not a hiring cycle.
The assessment step matters more than it sounds. Identifying the highest-return starting point means the first deployment pays for itself before expansion begins. That is the inverse of the CapEx model, where the full system goes in before any of it has been validated against your actual volumes. Best-practice implementation for warehouse robotics begins with baselining current metrics and defining clear use cases — the assess phase does exactly that, but produces a deployment instead of a planning document.
Here is the spec sheet ops teams actually plan against:
Throughput | 150 cases per hour |
Runtime | 16 hours per charge |
Payload | Cases up to 65 lb |
Upfront cost | $0 |
The throughput and runtime figures tell operators what a single unit can carry in a shift. The payload figure covers the vast majority of mixed-SKU boxed goods that move through distribution and wholesale operations. The upfront cost figure is the one that changes the business case entirely.
Proof points, not promises
The number that matters most on a spec sheet is the last one. Every incumbent in this space sells robots as equipment — capital expenditure, facility retrofitting, and a months-long integration. Lumper flips that denominator: you plan volume against a cost per pick, the same way you would price a temp worker, and scale up or down as orders move.
Warehouse robot cost is traditionally evaluated as a total-cost-of-ownership calculation spread over years. Industry guides on ROI for warehouse robotics quote payback periods of 8 to 24 months following the upfront capital outlay — and that range assumes the integration goes to plan, the volumes hold, and the facility does not need reconfiguration. For operators who cannot absorb a six-figure commitment to test a use case, those payback periods are not a path to automation. They are a reason to stay manual.
The pay-per-pick model converts warehouse robot cost from a fixed capital question into a variable operating line. The cost scales with the volume of work performed. When a shift spikes, you add robotic capacity instead of entering a hiring race you cannot reliably win. When volume falls, that capacity follows it back down — no idle six-figure machinery parked on your floor.
Case picking automation at scale has demonstrated productivity increases of up to 300% and labor cost reductions of up to 50% for operations that have made the CapEx commitment. The pay-per-pick model makes those productivity gains accessible without requiring the capital commitment as an entry condition. The gains are realized from the first pick, and the cost structure matches the output.
The operational benefit extends beyond cost. Operators who manage peak seasons against an unreliable temp labor supply face quality degradation as well as staffing gaps — pick accuracy falls when workers are unfamiliar with the floor, recently hired, or stretched across shifts they were not scheduled for. Robotic systems maintain consistent pick accuracy regardless of shift duration or volume level. That is a quality-of-output argument as much as a cost argument, and it compounds over peak periods.
What operators should do right now
Whether or not you talk to us, there is a smarter way to evaluate warehouse robotics than "how much does the robot cost":
Price it per pick, not per project. If a vendor can only quote you a CapEx number, you are not buying labor — you are buying a machine and betting volumes stay flat.
Check the deployment clock. "Six to eighteen months" should be a red flag, not a norm. If the robots cannot work in your existing layout, you are buying a construction project.
Start with the job that hurts most. You do not need to automate the whole floor at once. The shift you cannot keep staffed is the one that pays for automation first — start there.
The third criterion deserves emphasis. The impulse when evaluating automation is often to scope the full facility and price the full transformation. That approach is what produces the capital committee decision and the 18-month timeline. Starting with a single high-pain role — the dock job with the highest injury and turnover rate, the picking shift that temp agencies cannot reliably fill during peak — produces a bounded, measurable deployment with a clear return. It also produces operational experience with the system before committing to broader rollout.
WMS compatibility is the practical checkpoint that follows from criterion two. When a vendor quotes a six-month integration timeline, the underlying reason is almost always that the robotic system cannot consume the pick list your WMS already produces without a significant middleware project. That project has its own cost, its own risk, and its own failure modes. The right question to ask before any robotics conversation goes further is: what does your system need from my WMS to start a pick, and how long does that handoff take to configure?
What's next
Our first fleets are deploying into distribution and wholesale operations that run on manual labor today. Over the next 12 months our focus is straightforward: prove that autonomous warehouse robots can be hired as casually as a temp worker, and make "we're not ready for automation" a sentence no operator ever has to say again.
The 90% of U.S. warehouses that have no automation today did not stay manual because they failed to recognize the value. They stayed manual because every version of automation on offer was priced for a different kind of business. The CapEx model assumed stable volumes, available capital, tolerance for a months-long transition, and a facility that could absorb retrofitting. Most distribution and wholesale operations do not have all four of those conditions at once, and waiting until they do means leaving the labor problem unsolved indefinitely.
The pay-per-pick model removes each of those preconditions. There is no capital commitment, so stable volumes are not required to justify the investment. There is no facility retrofit, so the transition does not disrupt ongoing operations. There is no extended integration project, so the deployment clock is measured in hours rather than quarters. And because cost scales with output, operators can start with a single use case, validate the return, and expand from a position of demonstrated evidence rather than projected ROI.
Global warehouse automation spending exceeds $30 billion and continues to grow. The direction of the industry is not in question. What Lumper changes is which operators can participate in that shift, and on what terms.
If your floor moves cases, we would like to show you the per-pick numbers. Book a deployment call and tell us your volumes.
About Lumper
Lumper deploys autonomous warehouse robots for case picking, trailer loading, and palletizing — priced by the pick, with no upfront cost, no retrofitting, and no long integration. Built for the 90% of US warehouses that have no automation because existing robotics are too expensive and hard to deploy.
