Robot Leasing vs Buying: ROI Calculator & Decision Guide

The decision framework

Before the math, get the framing right. There are five variables that determine the lease-vs-buy answer for research robots:

1. Project length. A 12-month grant points to leasing. A multi-year program points to buying.
2. Utilization rate. If the robot will run 30+ hours per week, buying amortizes well. If it will sit idle much of the time, leasing is cheaper per productive hour.
3. Technology obsolescence risk. Humanoids and dexterous hands depreciate fast. Industrial arms (Franka, UR) depreciate slowly. Fast depreciation favors leasing.
4. Residual value risk. Do you believe you can resell this unit at a reasonable price in 24 months? If unsure, leasing transfers that risk to the lessor.
5. Tax and accounting treatment. Opex vs capex, Section 179, depreciation schedule. Details in the tax section below.

Worked example: Unitree G1, 24 months

Take a standard research deployment of a Unitree G1 for a 24-month research project.

Buy outright

Lease at $700/month with lease-to-own

The mid case puts buying and leasing within a thousand dollars of each other. The deciding factor is risk: leasing transfers the resale risk to the lessor. If you are confident the G1 will retain value and you want to own the asset, buy. If you are not, lease.

Simple ROI calculator (fill in your own numbers)

For an H1 (USD 90k base) over a 36-month horizon at a 4%/month lease rate, total lease payments are USD 129,600 and buying with a 50% residual lands near USD 81,000 net. For predictable long projects, buying wins by a meaningful margin. For short projects or uncertain utilization, leasing wins.

Tax treatment — US research and commercial entities

In the United States:

• Section 179 allows many small and mid-sized businesses to expense qualifying equipment purchases in the first year rather than depreciating over time. The 2026 Section 179 cap is in the multi-million range, well above any single-robot purchase. For research robots bought outright, this can be a substantial year-one tax benefit.
• Operating lease payments are typically deducted as current-year operating expenses, which is straightforward bookkeeping but spreads the deduction over the lease term.
• Capital lease treatment is different: the robot is booked as an asset and the lease as a liability; the lessee depreciates the robot. In practice SVRC’s standard lease is an operating lease unless specifically structured otherwise.
• Depreciation schedule for research robots under MACRS is typically 5 years, though the economic useful life is often shorter (3 years) for humanoids.
• University research accounts and grant-funded purchases have additional rules that vary by institution. Some grants forbid lease payments; others prefer them.

This guide is not tax advice. Consult a CPA before relying on any specific tax treatment.

Residual value and obsolescence risk

The single biggest financial risk in buying a research humanoid today is that a new, cheaper, better model ships six months later. In a maturing category this would not matter much; in the humanoid category in 2026 it matters a lot. A Unitree G1 bought in 2024 for USD 16,000 is worth roughly USD 6,000–$8,000 on the used market in early 2026, a 50–60 percent drop over two years. That is closer to consumer electronics depreciation than to industrial-arm depreciation.

Industrial arms are different. A Franka FR3 or UR5e bought in 2023 still holds 60–80 percent of its original value in 2026 on the secondary market, closer to a machine-tool depreciation schedule. For industrial arms, buying outright makes sense for most labs. For humanoids and dexterous hands, leasing is often the better hedge.

See our research humanoid cost guide for detailed pricing and our open-source vs commercial TCO guide for a platform-by-platform comparison.

Utilization rate: the unmeasured lever

Most labs under-estimate how often a research robot actually runs. A careful audit of 2024 lab utilization across SVRC customer sites found that typical research humanoids run 8–15 hours of active use per week, which is much lower than the 40-hour week most buyers plan for. At low utilization, the buy-vs-lease math tilts harder toward leasing because the per-productive-hour cost of a leased robot is only modestly higher than an owned one while you avoid the residual-value risk entirely.

Build a utilization estimate before committing to a purchase. If you cannot name the specific research tasks and operators who will fill 20 hours of weekly use, start with a lease.

When to buy (the short list)

• Long project (24+ months) with predictable utilization
• Platform is in a slow-depreciation category (industrial arm, quadruped)
• Grant restrictions favor capital purchases
• You want full control over modifications and firmware
• Institution policy or finance office strongly prefers capex

When to lease (the short list)

• Short project (6–18 months)
• Evaluating multiple platforms before committing
• High obsolescence risk (humanoids, dexterous hands, bleeding-edge sensors)
• Uncertain utilization
• Prefer opex treatment, maintenance bundled, predictable monthly cost
• Want optionality: return, extend, or buy out at term end

Three worked scenarios from real SVRC customers

Scenario 1: Stanford lab, G1 for grant-funded VLA research (18 months)

A lab receives an 18-month NSF grant to develop a humanoid VLA policy. Utilization is uncertain but likely low in months 1–3 (setup), peaking in months 9–15, declining at project end. Tech obsolescence risk is high. Recommendation: lease. An 18-month SVRC lease at USD 700/month totals USD 12,600, includes maintenance, and matches the grant lifecycle. Return the unit at project end and reclaim the lab space.

Scenario 2: Robotics startup, Franka FR3 for production data collection (36 months)

A startup building a proprietary VLA intends to collect data five days a week for at least three years. Utilization will be high (30+ hours/week). Obsolescence risk is lower (Franka is mature). Recommendation: buy. The USD 30,000 Franka plus USD 9,000 in support contracts over 36 months totals USD 39,000 vs approximately USD 54,000 for a 36-month lease at 5 percent/month. Buy outright, claim Section 179 (consult a CPA), and depreciate across the intended 3-year useful life.

Scenario 3: University teaching lab, SO-100 fleet for students (indefinite)

A department wants 10 SO-100 leader-follower pairs for a graduate manipulation course. Cost per pair is under USD 500. Leasing 10 open-source pairs makes no sense; there is no depreciation risk and no maintenance contract to bundle. Recommendation: buy outright. For low-BOM open-source arms, always buy.

Subscription, rental and event formats

Not every transaction is a clean lease or purchase. Three SVRC formats that often slot between them:

• Monthly rental. Short-horizon commitments, no lease paperwork. Used by labs doing 1–3 month pilots before committing to a lease.
• Event rental. Per-day or per-week pricing for demos, trade shows and corporate events. Starts around USD 500/day depending on platform.
• Subscription bundle. Robot + data services + support in a single monthly line item. Useful for buyers who want a turnkey arrangement rather than an equipment lease.

See event robotics for event formats and the main leasing page for standard lease inventory.

SVRC lease terms at a glance

Standard SVRC research-robot leases:

• Minimum term: 1 month (event/demo rentals as short as 1 day)
• Typical term: 3, 6, 12, 24 or 36 months
• Rate: 3–5% of purchase price per month for research robots; volume and multi-year discounts available
• Included: hardware, setup, documentation, data platform access; maintenance optional
• End of term: return, extend, upgrade, or convert to purchase with credit for a portion of prior payments
• Coverage: nationwide US shipping; same-day pickup at our Palo Alto location

See the full leasing page for current inventory and terms. Ask about lease-to-own credit structures when you request a quote.

Frequently asked questions

Next steps