How to Evaluate Drone Contract Manufacturers: Capacity, Compliance, and Production Readiness Criteria

Why Rigorous CM Evaluation Matters Now

The drone manufacturing landscape is undergoing rapid consolidation and expansion simultaneously. The DoD's Drone Dominance initiative is projecting a need for over 200,000 drones by 2027, a figure that will strain every link in the domestic supply chain. Meanwhile, the Blue UAS (Unmanned Aircraft Systems) program demonstrated just how few manufacturers can meet federal requirements: out of more than 300 companies that applied, only 23 made the cut. That is a pass rate below 8 percent.

For procurement managers and supply chain leads, the implication is direct. The pool of contract manufacturers that can actually deliver compliant, production-quality drone assemblies at scale is far smaller than the pool of companies claiming they can. A structured evaluation process is not bureaucratic overhead. It is the difference between a program that delivers on schedule and one that stalls at low-rate initial production for a year while the CM scrambles to stand up processes that should have been in place before the contract was signed.

This guide is not an RFQ template. It is not about preparing your requirements document. It is about evaluating the manufacturer itself: their facility, their systems, their compliance posture, their production capacity, and their ability to scale. A procurement manager should be able to use this framework to walk into a CM site visit, ask the right questions, and leave with a defensible assessment of whether that partner can execute.

Manufacturing Capability and Scalability Assessment

Current Production Capacity

Start with what the CM is actually producing today, not what their facility could theoretically support. Ask for current monthly unit output, current product mix, and capacity utilization as a percentage. A CM running at 90 percent utilization on existing contracts has no room for your program without capital investment and hiring, both of which introduce schedule risk.

Evaluate the production floor directly. Look for dedicated assembly areas with ESD controls, conformal coating capability, clean benches for optics and sensor integration, and functional test stations with documented procedures. A CM that builds drones on the same bench where they rework telecom boards is not a serious candidate for defense or commercial UAV programs.

Surge Capacity

Surge capacity is the ability to increase output in weeks, not months. In the current demand environment, with DoD programs accelerating timelines and commercial operators placing larger orders, surge is not a nice-to-have. It is a contract requirement.

Ask the CM to describe their surge plan in concrete terms. How many additional shifts can they staff? What is the lead time on additional test equipment? Do they have pre-qualified subcontractors they can activate? A credible surge plan includes specific numbers: "We can go from 50 units per month to 150 units per month within six weeks by adding a second shift and activating our overflow assembly partner in [location]." A vague answer like "we can scale as needed" is a red flag.

Assembly Partner Onboarding

If the CM uses or plans to use assembly subcontractors, evaluate the onboarding process. With good infrastructure, meaning documented work instructions, fixture kits, test procedures, and training materials, a competent assembly partner can be onboarded in 2 to 4 weeks. If the CM quotes 3 to 6 months to bring up a new line, their documentation and process maturity are insufficient for a program that may need to scale rapidly.

The onboarding timeline is a proxy metric. It reveals how well the CM has codified their manufacturing knowledge. Companies that rely on tribal knowledge held by a few experienced technicians cannot scale, cannot surge, and cannot recover from the departure of key personnel.

Quality Systems and Certifications

AS9100D Certification

AS9100D is the aerospace quality management standard, and it is the baseline expectation for any CM serving defense or aviation-adjacent programs. Do not accept a CM's claim of AS9100D certification at face value. Verify it through the IAQG OASIS database at iaqg.org/oasis, which is the authoritative registry for aerospace quality certifications worldwide. The database lists the CM's certification scope, their registrar, and the certification expiration date.

Pay attention to the scope. A CM may hold AS9100D for "electronic assembly" but not for "mechanical assembly and integration." If your drone program requires structural assembly, harness fabrication, or system-level integration, the certification scope must cover those activities. A scope mismatch means the CM's quality system has not been audited against the processes your program actually requires.

NADCAP Accreditation for Special Processes

If your drone design involves special processes such as soldering to IPC J-STD-001 Class 3, conformal coating, composite layup, heat treatment, or non-destructive testing, the CM should hold NADCAP (National Aerospace and Defense Contractors Accreditation Program) accreditation for those specific processes. NADCAP is a third-party audit program that goes deeper than AS9100D into the technical execution of individual manufacturing processes.

Not every drone CM will need NADCAP. But if the design involves any special process, and the CM does not hold the corresponding NADCAP accreditation, you need to understand how they control that process and whether their customer base has accepted their approach. For defense programs, the absence of NADCAP for a required special process is typically a finding during government quality audits.

First Article Inspection: AS9102

First article inspection (FAI) per AS9102 is not optional. It is the formal gate that proves a CM can produce a conforming part or assembly before production begins. The FAI package documents every dimension, every material, every process, and every test result for the first production unit, traced back to the engineering drawing and specification.

Evaluate the CM's FAI process by asking to see a completed FAI package from a previous program (with proprietary data redacted). What you are looking for: completeness, traceability to drawing revisions, clear identification of any non-conformances and their dispositions, and evidence that the FAI was reviewed and approved by quality engineering, not just rubber-stamped by a technician. A CM that treats FAI as a paperwork exercise rather than a genuine manufacturing verification gate will produce the same attitude toward production quality.

Defect Tracking and Root Cause Resolution

Ask to see the CM's corrective action system. Specifically, ask for metrics: What is their defect rate by product line? What is their average time to root cause? What is their corrective action closure rate? How do they track recurrence?

A mature CM will have a formal corrective and preventive action (CAPA) system, typically integrated into their quality management system, with defined escalation paths, root cause analysis methodologies (8D, fishbone, 5-why), and evidence of effectiveness verification. A CM that cannot produce defect trend data on request either does not track defects systematically or does not want you to see the numbers. Neither is acceptable.

NDAA Compliance and Supply Chain Security

Understanding Section 848 Requirements

NDAA Section 848 identifies specific drone component categories that are subject to sourcing restrictions for federal procurement. The covered components list includes: flight controllers, radios and communication systems, cameras, gimbals, ground control stations (GCS), batteries, motors, and navigation systems. Any CM building drones for government end use must demonstrate that none of these components originate from covered entities.

This is not a checkbox exercise. It requires the CM to maintain detailed supply chain documentation that traces every covered component back to its manufacturer, its country of origin, and its compliance status. The CM should be able to produce this documentation on demand, not after a two-week scramble through their purchasing records.

Approved Vendor List Discipline

An Approved Vendor List (AVL) is only as good as the process that governs it. Evaluate the CM's AVL discipline by asking these questions:

• How are new vendors qualified and added to the AVL?
• What triggers a vendor re-evaluation or removal?
• How frequently is the AVL reviewed against current NDAA and ITAR requirements?
• Can the CM demonstrate that every component on your bill of materials maps to an AVL-approved source?
• What is the process when a preferred vendor has a supply disruption? Is there a pre-qualified alternate?

Strong AVL discipline means the CM has pre-qualified alternate sources for critical components, maintains incoming inspection procedures that verify components match AVL specifications, and has a formal process for managing vendor deviations. Weak AVL discipline means the buyer discovers six months into production that a critical NDAA-covered component was substituted with an unapproved part because the original went on allocation.

Lot-Level Component Traceability

Lot-level traceability means the CM can identify, for any finished drone, which specific lot of each component was used in its assembly. This is not the same as knowing which vendor supplied the component. It is knowing that the flight controller in serial number UAV-2026-0847 came from lot FC-2026-0312, which was received on a specific date, passed incoming inspection per a specific procedure, and was stored under controlled conditions until use.

This level of traceability is essential for two reasons. First, if a component defect is discovered in the field, lot-level traceability enables targeted containment rather than a fleet-wide recall. Second, for NDAA compliance, lot-level traceability provides the evidence chain that proves component provenance for every unit delivered, not just at the time of vendor qualification.

Supply Chain Documentation Requirements

At minimum, the CM should maintain and be able to produce the following supply chain documentation for every NDAA-covered component:

• Certificates of conformance (C of C) from the component manufacturer
• Country of origin declarations
• Incoming inspection records with accept/reject disposition
• Storage and handling records demonstrating environmental controls (especially for batteries and optics)
• DFARS compliance documentation where applicable
• End-user certificates for ITAR-controlled items

If the CM cannot produce this documentation stack in an organized, retrievable format, their supply chain management system is not mature enough for government drone programs. Period.

Hardware Quality Assessment

Beyond the management system, evaluate the physical quality of what the CM actually produces. Request sample units or visit the production floor during active assembly and test.

Build Durability

Inspect the mechanical assembly for evidence of proper fastener torque (torque-stripe marking or documented torque values), adequate wire routing and strain relief, conformal coating coverage on exposed electronics, proper sealing on weather-rated enclosures, and robust mounting of vibration-sensitive components like cameras and IMUs. A drone that looks clean on the bench but sheds fasteners after 20 flight hours is a warranty liability and a safety risk.

Payload Capacity and Flight Time

If the CM is building to your design, verify that their production units meet the payload and endurance specifications from the engineering baseline. Ask for flight test data from production units, not prototypes. Production units often weigh more than prototypes due to conformal coating, production-grade connectors, and properly routed harnesses. A 5 to 10 percent weight increase from prototype to production is normal; a 20 percent increase indicates the prototype was not representative of producible design.

Sensor Ecosystem Compatibility

Evaluate the CM's experience integrating the sensor payloads your program requires: EO/IR cameras, LiDAR, multispectral sensors, SAR, or other mission-specific payloads. The CM should demonstrate familiarity with the mechanical, electrical, and data interfaces for your sensor suite. Ask how they handle sensor calibration during production, whether they have fixtures for optical alignment, and whether their test procedures include end-to-end sensor verification, not just a power-on check.

Unit-Level Traceability and Production Records

Every finished drone that leaves the CM's facility should carry a complete unit-level data package. This is the manufacturing equivalent of a vehicle's title and service history. The data package should include:

• Serial number of the completed system and every serialized subassembly
• Lot numbers for every component installed
• Firmware versions loaded at the time of delivery, including flight controller, ESC, radio, and GCS software
• QA sign-off at each inspection gate, with inspector identification
• Functional test results including motor run-up, communication range check, sensor verification, and pre-delivery flight test (if applicable)
• Non-conformance records for any deviations encountered and their disposition
• Calibration records for any test equipment used during final acceptance

This unit-level traceability package serves multiple purposes. It enables field support teams to diagnose issues with knowledge of exactly what is in each airframe. It provides the evidence chain for regulatory compliance. And it gives the program office the data needed to identify trends across the fleet, catching a systematic issue at unit 50 rather than unit 500.

Ask the CM to show you a completed unit-level data package from a current production program. If they cannot produce one, or if the package is incomplete, their production documentation system is not ready for your program.

Financial Stability Assessment

A CM that delivers excellent quality but goes insolvent during your production run is worse than a CM you never selected. Financial stability assessment is an uncomfortable but necessary part of the evaluation.

For publicly traded CMs, financial data is accessible through SEC filings. For private companies, request audited financial statements or, at minimum, a D&B (Dun & Bradstreet) report. Key indicators to evaluate:

• Revenue concentration: If more than 40 percent of the CM's revenue comes from a single customer, the loss of that customer could destabilize the business.
• Debt-to-equity ratio: Heavily leveraged CMs may struggle to fund the capital expenditure required to scale for your program.
• Cash reserves: Manufacturing is capital-intensive. A CM with thin cash reserves may pressure you for accelerated payment terms or advance payments to fund material purchases.
• Backlog: A healthy backlog indicates market confidence. An empty backlog could mean the CM is desperate for work and may underbid to win your contract, leading to quality and delivery problems later.
• Insurance: Verify adequate product liability coverage, especially for drones operating in the national airspace or in proximity to people.

Financial assessment is not about finding a rich CM. It is about ensuring the CM you select will still be operating, investing in their facility, and retaining their workforce for the duration of your program, which for defense contracts may be 5 to 10 years including sustainment.

CM Evaluation Scorecard

The following scorecard provides a structured framework for comparing contract manufacturers. Score each criterion from 1 (does not meet) to 5 (exceeds expectations). A minimum score of 3 in every category should be required for consideration, with any score of 1 in a critical category being disqualifying.

Red Flags That Disqualify a Manufacturer

Certain findings during evaluation should be treated as disqualifying, not as areas for improvement. These are not weaknesses to manage. They are indicators that the CM is not ready to support a serious drone production program.

No AS9100D Certification or Expired Certification

If the CM claims aerospace capability but does not hold current AS9100D certification, or the certification has lapsed, they have either failed to maintain their quality system or never invested in one. "We are pursuing certification" is not the same as holding it. Certification takes 12 to 18 months and a significant organizational commitment. Walk away and revisit when they have the certificate in hand and verified in the IAQG OASIS database.

Cannot Demonstrate NDAA Compliance for Covered Components

If the CM cannot produce documentation showing compliant sourcing for every Section 848 covered component category, they are not ready for government work. This is not a documentation gap. It is a supply chain gap. Remediating non-compliant sourcing for flight controllers, radios, cameras, gimbals, GCS, batteries, motors, and navigation systems can take 6 to 12 months of vendor qualification and redesign.

No First Article Inspection Process

A CM that does not perform formal first article inspection per AS9102, or cannot show completed FAI packages from prior programs, has no mechanism for verifying that their production process actually produces conforming product. Without FAI, the first production run is an experiment, and your program is the test subject.

Tribal Knowledge Instead of Documentation

If the site visit reveals that critical assembly steps are known only to specific technicians, if work instructions are informal or missing, if test procedures are not documented, the CM cannot scale, cannot onboard assembly partners, and cannot survive the departure of key personnel. This is the most common failure mode in small and mid-size drone CMs, and it is the hardest to remediate because the people who hold the knowledge often do not recognize the problem.

Inability to Provide Unit-Level Traceability

If the CM cannot produce a complete traceability record, including serial, lot, firmware, QA sign-off, and test results, for units currently in production, they will not be able to produce one for your program. Promises to implement traceability systems "before your program starts" should be treated with skepticism. These systems require months to implement and validate.

Financial Instability or Excessive Customer Concentration

A CM that refuses to provide financial information, or whose financials reveal critical vulnerabilities such as negative cash flow, excessive debt, or dependence on a single customer for more than half their revenue, poses an unacceptable program risk. The cost of a CM going insolvent mid-production far exceeds the cost of selecting a slightly more expensive but financially stable partner.

No Corrective Action Data

A CM that cannot produce defect rate data, corrective action records, or root cause analysis examples either does not track quality performance or is hiding poor performance. Both are disqualifying. A CM that tracks defects and has a 3 percent defect rate with downward trend is a better partner than a CM that claims zero defects because they do not measure.

Vague Surge Capacity Claims

When asked about surge capacity, if the CM responds with generalities rather than a specific plan with identified resources, timelines, and constraints, they have not thought about scaling. In a demand environment where the DoD is projecting 200,000-plus drones by 2027, a CM that cannot articulate how they would double or triple output is not positioned for the market they claim to serve.

Structuring the Evaluation Process

Phase 1: Desktop Review

Before scheduling a site visit, conduct a desktop review. Verify AS9100D certification in the IAQG OASIS database. Check for NADCAP accreditations. Review the CM's public-facing information for consistency with their claimed capabilities. Request and review their capability brief, quality manual summary, and references from current customers. This phase should eliminate 40 to 60 percent of candidates and takes one to two weeks.

Phase 2: Facility Audit and Technical Evaluation

For CMs that pass the desktop review, conduct an on-site audit. Walk the production floor. Review quality records. Inspect work in progress. Interview quality engineers, production supervisors, and supply chain managers. Use the evaluation scorecard to document findings in real time. This phase takes one to two days per CM.

Phase 3: Pilot Build

Before committing to a production contract, require a pilot build of 5 to 10 units. The pilot build validates the CM's ability to execute your specific program, not just their general capability. Evaluate the pilot build for workmanship quality, dimensional conformance, test results, documentation completeness, and on-time delivery. The pilot build should include a complete first article inspection per AS9102.

Phase 4: Production Contract with Performance Gates

Structure the production contract with clear performance gates tied to quality metrics, delivery schedule, and cost targets. Include provisions for periodic quality audits, access to production data, and defined escalation procedures for non-conformances. Build in surge clauses that define the CM's obligations and the timeline for capacity increases.

Building a Defensible Selection Decision

The evaluation framework described here is deliberately thorough. Every criterion exists because real programs have failed at that specific point. CMs have lost AS9100D certification mid-contract. NDAA non-compliant components have been discovered in delivered units. CMs have gone insolvent, leaving programs without a manufacturing partner. Production quality has collapsed because the only technician who knew how to align a sensor payload left the company.

The Blue UAS program's 23-out-of-300 pass rate was not the result of arbitrary standards. It was the result of applying exactly the kind of evaluation criteria described here: manufacturing capability, compliance posture, quality systems, and production readiness. The companies that passed had invested years in building the systems and infrastructure that make reliable, compliant, scalable drone manufacturing possible.

For procurement managers evaluating drone contract manufacturers, the goal is not to find a perfect CM. It is to make a defensible selection based on evidence, not promises. Document your evaluation. Score it consistently. Verify claims independently. And treat red flags as what they are: indicators of risk that no amount of contract language will mitigate.

The drones your program fields will only be as reliable as the manufacturer that built them. Invest the time in getting the selection right.