National Key R&D Program Project Completion Testing: Scope, Standards, and Acceptance Verification

National Key R&D Program project completion testing is the mandatory third-party verification process required before research deliverables can pass the final acceptance review administered by the Ministry of Science and Technology of China. Accredited third-party laboratories perform functional, performance, reliability, and indicator-specific tests to confirm that prototypes, software systems, materials, and integrated platforms meet the key technical targets defined in the project task book. This article explains the scope, sample categories, testing items, methods, applicable standards, and frequently asked questions related to the completion testing process.

Technical Overview

The National Key R&D Program is the flagship government-funded research initiative of China, established in 2016 following the national science and technology management reform. The program is administered by the Ministry of Science and Technology (MOST) and implemented through a series of special projects covering strategic fields such as advanced manufacturing, new energy, information technology, agriculture, ecological environment, public safety, and deep-sea, space, and polar exploration. Each funded project is governed by a formally approved task book that defines research objectives, deliverables, and quantified key technical indicators.

Project completion testing, also referred to as project conclusion testing or acceptance testing, is the independent verification activity performed at the end of the project implementation period. Under the interim management measures of the National Key R&D Program, most projects must submit third-party test reports as prerequisite evidence for the final acceptance review conducted by MOST or its authorized professional management institutions. The purpose of completion testing is to objectively confirm that the developed prototypes, software systems, materials, and demonstration platforms have achieved the technical indicators committed in the task book.

Testing must be carried out by accredited third-party laboratories holding CMA (China Metrology Accreditation) qualifications and preferably CNAS accreditation in accordance with ISO/IEC 17025, with the testing scope explicitly covering the standards and product categories involved. Test results documented in a formal report serve as core supporting evidence during acceptance review, follow-on funding evaluation, intellectual property substantiation, and subsequent product certification or industrialization activities.

Testing Samples

Hardware Prototypes and Equipment

Typical hardware samples include newly developed prototypes, pilot-scale equipment, precision instruments, sensors, robotic systems, power electronic devices, communication terminals, and medical or agricultural machinery. These samples are tested in their configured states as described in the project deliverable list, and their firmware versions, configuration parameters, and operating conditions are recorded to ensure traceability.

Software Systems and Platforms

Software deliverables include research application software, data processing platforms, industrial control software, simulation systems, database systems, and algorithm modules. Testing is performed on deployed, runnable builds with defined hardware environments, and both functional behavior and quality characteristics are evaluated against the requirements specification.

Materials and Components

Advanced materials developed under the program, such as alloys, composites, ceramics, polymers, coatings, catalysts, and electronic components, are submitted in the form of machined specimens or finished parts. Sampling plans are established according to relevant material testing standards to ensure that the characterized properties statistically represent the material system achieved by the project.

Integrated Systems and Demonstration Engineering

Large-scale, immovable, or site-dependent deliverables, such as demonstration production lines, environmental monitoring networks, energy stations, and infrastructure monitoring systems, are typically verified through on-site testing at the demonstration location, supplemented by laboratory testing of detachable subsystems and modules.

Testing Items

Completion testing is organized around two complementary objectives: verification of the task book indicators and verification of general quality characteristics. Common testing items include the following:

• Key technical indicator verification: quantified measurement of the specific performance targets committed in the task book, such as precision, throughput, detection limits, energy efficiency, response time, capacity, or yield, using calibrated instrumentation and documented test methods.
• Functional testing: systematic verification that each documented function of hardware or software deliverables operates as specified, including normal-path, boundary, and exception-handling behavior.
• Performance testing: evaluation of processing speed, latency, concurrency, data volume handling, dynamic response, and resource utilization under representative and worst-case load conditions.
• Reliability testing: assessment of mean time between failures (MTBF), mean time to repair (MTTR), duty cycle endurance, continuous operation stability, and stress retention, planned in accordance with reliability engineering methods such as those in the GB/T 5080 series.
• Environmental adaptability testing: verification of operating, storage, and transport performance under temperature, humidity, vibration, shock, salt spray, and IP ingress conditions following the GB/T 2423 series, which aligns with IEC 60068.
• Electrical safety testing: insulation resistance, dielectric strength, leakage current, grounding continuity, and protective circuit verification in line with GB 4943.1, which corresponds to IEC 62368-1, or applicable product-specific safety standards.
• Electromagnetic compatibility (EMC) testing: conducted and radiated emissions and immunity testing following the GB/T 17626 series and related IEC 61000-4 based methods where relevant to the deliverable class.
• Software quality testing: evaluation of functional suitability, performance efficiency, compatibility, usability, reliability, security, maintainability, and portability under the quality model of ISO/IEC 25010 and the requirements of GB/T 25000.51.
• Information security testing: vulnerability scanning, access control verification, audit logging review, and compliance assessment against GB/T 22239 for systems subject to cybersecurity classified protection requirements.
• Accuracy and metrology verification: calibration checks, measurement error analysis, repeatability and reproducibility studies, and uncertainty evaluation for instruments and measurement methods developed by the project.

Testing Methods

Indicator Decomposition and Test Planning

The testing process begins with a detailed analysis of the project task book and acceptance criteria. Each key technical indicator is decomposed into verifiable test requirements, and an indicator-to-test traceability matrix is established. A formal test plan is then prepared, defining test objectives, sample configuration, environmental conditions, instrumentation, acceptance thresholds, and pass criteria, and is confirmed with the project team before execution.

Standards-Based Test Design

Wherever applicable, testing is conducted in accordance with national, international, and industry standards, including GB and GB/T standards, ISO and IEC international standards, ASTM methods for materials characterization, EN standards for equipment categories, UL and JIS standards where product destinations require them, and documented project-specific methods where no published standard exists. Custom test methods are validated through repeatability trials, reference comparisons, and uncertainty assessment before use.

Laboratory and On-Site Execution

Laboratory testing is performed within accredited facilities using calibrated equipment under controlled environmental conditions. For large or site-bound deliverables, on-site testing is conducted at demonstration locations, with portable instrumentation, witnessed test procedures, and recorded raw data. Software testing combines black-box functional testing, automated regression suites, load generation tools, and static or dynamic code analysis as appropriate to the risk level and documentation available.

Data Recording, Review, and Reporting

All measurements are recorded with instrument identification, calibration status, and environmental data. Results are compared against the acceptance thresholds defined in the test plan, and deviations are documented and analyzed. The final test report states the testing basis, sample description, test items, methods and standards applied, test conditions, results, and conclusions for each task book indicator. Reports issued with the CMA mark provide the evidentiary weight required for acceptance submission.

Applications

Completion testing services support National Key R&D Program participants across the full spectrum of special projects, including intelligent manufacturing and robotics, advanced structural and functional materials, new energy and energy storage, next-generation information and communication technology, deep-sea, space, and polar exploration, agricultural technology, ecological environment governance, food safety, public security, disaster prevention, and biomedicine. Beyond acceptance review, the resulting test reports are widely used for:

• Final acceptance and performance evaluation by MOST professional management institutions and expert panels.
• Follow-on project application and continued funding justification.
• Evidence supporting patent applications, software copyright registration, and scientific achievement awards.
• Transition of research outcomes toward compulsory product certification, voluntary certification, or market entry under subsequent commercialization.
• Technical due diligence in technology transfer, licensing negotiations, and industry demonstration programs.

FAQ

What is the difference between completion testing and routine product testing?

Completion testing is acceptance-oriented rather than market-oriented. Its primary purpose is to independently verify the key technical indicators committed in the project task book, and the test scope, thresholds, and methods are derived from the project acceptance requirements rather than solely from commercial product standards.

Which qualifications should the testing laboratory hold?

The laboratory should hold CMA accreditation covering the relevant testing standards, and CNAS accreditation to ISO/IEC 17025 is strongly recommended. The laboratory must demonstrate independence from the project undertaking entities to preserve third-party status, and its accredited scope must explicitly cover the product categories and standards involved.

When should a project team engage the laboratory?

Early engagement is advisable, ideally during the mid-term phase. Early review of the task book indicators allows ambiguous or unverifiable indicators to be refined, test methods to be agreed in advance, and pre-test self-assessments to be conducted, which substantially reduces the risk of failed verification at the conclusion stage.

How long does completion testing take?

For software systems, testing typically requires one to four weeks depending on scale and defect density. For hardware involving environmental, reliability, or EMC sequences, four to ten weeks is common, and on-site demonstration system verification may extend further depending on scheduling and site readiness.

What happens if a key indicator fails to meet the task book target?

Failed items are documented with full test data and deviation analysis. The project team may perform corrective engineering work and request a retest of the affected items, or, where complete attainment is not feasible, disclose the deviation with justification in the acceptance materials in accordance with program management requirements.

Can testing be performed at the project site instead of a laboratory?

Yes. For immovable, large-scale, or geographically constrained demonstration systems, accredited laboratories can conduct on-site testing using transported instrumentation and witnessed procedures, provided that environmental conditions and measurement traceability requirements can be satisfied.