Compare robot actuator routes over a defined programme quantity using nonrecurring engineering, validation and cost per accepted actuator. Apply the same functional and integration envelope to both alternatives, then check capability, schedule and support before selecting one. In the fictional model below, bought actuators cost less for a five-robot pilot, while own-design actuators cost less at 100 robots if the stated assumptions hold.
The cost equality occurs at 160 actuators, or 80 robots using two each. That threshold is a result of this model's inputs. It supplies no evidence that a custom actuator has been developed, qualified or accepted for use.
Define the supply boundary and shared envelope
R-01's original fictional pilot uses two in-house DM-100 A drive modules per robot, each containing one purchased MT-110 A motor. The five units therefore need ten motors. That historic structure remains unchanged. This article evaluates a future complete-actuator route and creates no replacement approval for the original pilot. R-01 BOM structure
The future alternatives each supply a complete actuator containing the motion hardware, position sensing and local drive electronics defined by the candidate specification. One route buys that accepted assembly from a supplier. The other uses an own-design assembly, whether manufactured inside the company or by a contractor. Design authority and manufacturing location are separate decisions.
Both routes must satisfy the same candidate functional envelope and external interfaces. The following is an illustrative comparison basis, with values and requirement IDs to be replaced by the team's actual controlled brief.
| Shared comparison subject | Requirement or record to resolve for both routes | Responsible owner |
|---|---|---|
| Output behaviour | Same specified torque in N·m, speed in rpm, duty profile and load conditions | Motion owner |
| Mechanical installation | Same accepted mounting interface, envelope in mm and shaft or flange load limits | Mechanical owner |
| Electrical and control | Same input envelope in V/A, command/status contract and required fault behaviour | Electrical and controls owners |
| Thermal and life | Same installation temperature in °C, cooling conditions and defined life evidence | Thermal and verification owners |
| Product acceptance | Same applicable criteria, inspection scope and system-integration evidence | Verification and release authority |
| Programme supply | Quantities, dates, accepted source, replacement path and support scope | Buyer and programme owner |
Keep the physics and supplier-rating assessment with the actuator sourcing brief. Here, the financial comparison starts after the alternatives share a sufficiently defined boundary. If one quote excludes a gearbox, drive or required check, reconcile its scope before comparing it with a complete accepted actuator.
NASA's decision-analysis guidance distinguishes mandatory and enhancing criteria and requires consideration of assumptions and uncertainty. An option that fails a mandatory criterion is excluded from selection even if its cost is attractive. This is method support for the comparison, with no NASA obligation assigned to R-01. NASA decision-analysis guidance
Build a full programme cost model
Use two inputs for each route, fixed programme cost and recurring cost per accepted actuator. The fictional USD assumptions below include the planned bench and qualification work inside NRE. They represent budgets and unit estimates, with no actual quotation, test or delivered programme result.
| Cost input | Own-design route | Bought-equivalent route | Accounting rule |
|---|---|---|---|
| Nonrecurring engineering and integration | $40,000 | $5,000 | Charge once per route and defined programme scope. |
| Planned bench and qualification work | $8,000 | $3,000 | Included in NRE totals below, with no completed evidence implied. |
| Total NRE | $48,000 | $8,000 | Sum each route's two fixed lines once. |
| Recurring cost per accepted actuator | $220 | $470 | Includes the defined manufacturing or supply and unit-acceptance scope. |
| Actuators per robot | 2 | 2 | Use actuator quantity in the equations. |
The unit estimates assume production labour, purchased materials or supply price, manufacturing overhead and expected unit-level acceptance burden are accounted for. A real team needs cost evidence for those assumptions, including yield and rework. Whole-robot work common to both routes is outside this actuator model. Freight, tax, financing, warranty exposure and any additional tooling must be added or explicitly shown as common or excluded in the real comparison.
For actuator quantity N, the equations are C_make = 48,000 + 220N and C_buy = 8,000 + 470N. The model does not add a separate validation charge after NRE, because the $8,000 and $3,000 work budgets are already included.
Compare the pilot and production horizons
Five robots need ten actuators. A programme of 100 robots needs 200 actuators. The two rows below are alternative complete horizons for the same route assumptions. The 100-robot row is a total programme quantity, rather than 100 additional robots after the pilot.
| Total robots in the horizon | Actuators N | Own-design total | Bought total | Lower modelled total |
|---|---|---|---|---|
| 5 | 10 | $48,000 + $2,200 = $50,200 | $8,000 + $4,700 = $12,700 | Bought by $37,500 |
| 100 | 200 | $48,000 + $44,000 = $92,000 | $8,000 + $94,000 = $102,000 | Own-design by $10,000 |
If the same route later grows from five robots to 100, add the remaining 190 actuators to that route's recurring cost and keep its NRE charged once. Do not add both full-horizon totals. If the programme buys the pilot and later switches to own-design, prepare a separate transition model containing the actual sunk buying cost, new design NRE and remaining units.
Calculate equality and test a lower bought price
The fixed-cost difference is $48,000 − $8,000 = $40,000. The recurring saving from own-design is $470 − $220 = $250 per actuator. Equality is therefore N = 40,000 / 250 = 160, or 160 / 2 = 80 robots.
Both routes total $83,200 at N=160. Below 160 actuators the bought route has lower modelled total cost. Above 160, own-design has lower modelled total cost under these inputs. An 80-robot threshold should not be confused with proof of future demand or an engineering release.
A separately labelled sensitivity assumes a bought-equivalent quote of $420 per accepted actuator, with bought NRE still $8,000. The quote is fictional and preserves the same scope. It is an alternative scenario, never an additional charge on the original model.
| Sensitivity horizon | Own-design at $220/unit | Bought at $420/unit | Result |
|---|---|---|---|
| 10 actuators | $50,200 | $12,200 | Bought remains lower by $38,000. |
| 200 actuators | $92,000 | $92,000 | Equality at 100 robots. |
With the lower bought price, the recurring saving becomes $200 per actuator and equality moves to 40,000 / 200 = 200 actuators. A quote change of $50 per actuator therefore removes the original $10,000 own-design advantage at the 200-unit horizon.
For these higher own-design fixed costs, no positive cost crossing exists if the own-design recurring cost is equal to or higher than the bought recurring cost. A zero or negative per-unit saving cannot repay the extra $40,000 NRE. The general equation needs a positive denominator before it yields a meaningful crossing quantity.
Keep non-cost gates in the decision
The programme owner chooses the quantity horizon and acceptable exposure. Engineering determines whether each route meets the shared envelope. Purchasing checks the offer, source and delivery scope. Verification defines and accepts the required evidence.
| Gate | Evidence before selecting or releasing a route | Hold or action |
|---|---|---|
| Required capability | Complete offered or own-design actuator evidence against the controlled brief | Hold unsupported operating points or life claims. |
| Interfaces | Accepted mechanical, electrical and command contract with exact revisions | Review any difference and resulting integration cost. |
| Development and delivery dates | Staffed design/test plan or dated supplier commitment with receipt allowance | Hold schedule claims without an executable plan. |
| Validation and qualification | Scoped bench and system evidence plan, then actual reviewed results | NRE allowance does not close the engineering gate. |
| Alternative supply and support | Source history, replacement route, access and responsibilities | Record unresolved dependency and its programme consequence. |
| Demand and economics | Credible quantity horizon, comparable quote and sensitivity | Revisit the choice when assumptions could change the ranking. |
In the fictional low-volume decision, the team recommends the bought route for a five-robot future evaluation because its budget is lower. Production release remains conditional on the capability, interface, supply and evidence gates. The 100-robot own-design route stays an alternative to assess, with the $420 sensitivity explaining why a new quote could change the cost preference.
Record the actual recommendation, the authority's decision, conditions and actions in an engineering decision record. A contractor can manufacture a custom actuator while the robot team retains design approval. Price comparison does not settle that authority arrangement.
Robot actuator make-or-buy questions
Is the equality point 80 or 160?
It is 160 actuators, equivalent to 80 robots at two actuators each. Both routes total $83,200 there under the original model.
Should validation be added again after NRE?
The example's validation budgets are already inside its NRE. Add extra work only when it is outside that defined scope, and revise the model explicitly.
Does own-design mean manufacturing in our workshop?
The team can own the design and outsource manufacture. Record design authority, source selection, manufacturing responsibility and acceptance separately.
Can a cheaper bought quote change the choice?
Yes. At $420 per accepted actuator, the model reaches equality at 200 actuators. Reconcile the quote's technical and acceptance scope before using its price.
Continue the decision review
Use the actuator sourcing guide for the shared capability brief and the engineering decision record for authority, uncertainty and resulting actions.