NEWF/WORKS

SBIR / Prototyping and test

Turn a technical claim into
evidence you can act on.

For SBIR topics involving software or integrated systems, NewfWorks identifies the key technical risk, builds enough of the system to test it, and records the results of repeatable tests. Phase I ends with a clear technical conclusion, working code, and test results that can be extended in Phase II.

01ClaimWhat must be demonstrated?
02BuildWhat must be implemented to test it?
03EvidenceWhat result supports the decision?

01 / FRAME THE WORK

Test the assumption that could break the approach.

In Phase I, we isolate the technical assumption most likely to invalidate the concept, then build only enough of the end-to-end system to test it under defined, repeatable conditions.

TECHNICAL DECISION LOOP / 01 From technical question to decision
DEFINED CLAIM · REPEATABLE TEST · TRACEABLE RESULT
01 / CLAIM Testable claim use case · threshold 02 / BUILD Working prototype critical path · interfaces 03 / TEST Measured result baseline · candidate 04 / DECIDE Decision record result · uncertainty RESULTS REFINE THE CLAIM OR DEFINE THE NEXT BUILD
  1. 01 / CLAIMTestable claimUse case · success threshold
  2. 02 / BUILDWorking prototypeCritical path · interfaces
  3. 03 / TESTMeasured resultBaseline · candidate
  4. 04 / DECIDEDecision recordResult · remaining uncertainty
CLAIMUser · scenario · success threshold
PROTOTYPECritical path · working interfaces
TEST RECORDCode version · test setup · measured result

For each decision, we retain the code version, configuration, test conditions, and measured results.

01 / CLAIM

Define what must be true

State the user, operating scenario, inputs, outputs, performance threshold, and boundary conditions.

02 / BASELINE

Choose a credible comparison

Measure the current method or simplest alternative under the same data, load, environment, and scoring rules.

03 / DECISION

Set the decision rule before testing

Define success, failure, and stop criteria, including how uncertainty and invalid trials will be treated.

02 / PHASE CONTINUITY

Carry Phase I engineering into Phase II.

We keep source code, test code and data, interface definitions, and decision records versioned and usable. We also document deployment constraints so Phase II can extend the system instead of rebuilding it.

00

Topic

Frame the problem

Define the user, operating scenario, system boundary, key assumption, comparison baseline, and acceptance criteria.

01

Phase I

Test feasibility

Build the smallest end-to-end implementation needed to compare the approach against its baseline under controlled conditions.

02

Phase II

Integrate and harden

Replace test doubles with operational interfaces, then address performance, cybersecurity, data handling, deployment, and support.

03

Phase III

Transition and sustain

Move the technology toward operational use with non-SBIR funding, a designated integration owner, operating procedures, a support plan, and evidence that it meets acceptance criteria.

DELIVERED TOGETHER

Working system.
Traceable results.

We develop the prototype and its test package together, keeping each result traceable to the code version, configuration, data, and test conditions that produced it.

WORKING IMPLEMENTATION

Build the critical path

  • Software, models, devices, and operator actions connect through defined interfaces
  • The design accounts for limits on compute, timing, power, storage, and network capacity
  • The implementation enforces authority boundaries and defines behavior when part of the system fails
  • Shortcuts, test doubles, and external dependencies are documented
TEST PACKAGE

Make the results reproducible

  • Each test case corresponds to an operational scenario and reference dataset
  • Accuracy, uncertainty, latency, reliability, and cost are reported separately
  • The package records configuration, software versions, data provenance, and results
  • Tests cover nominal, boundary, and degraded conditions

03 / REUSABLE OUTPUTS

Engineering assets that carry into
the next phase.

A

Working end-to-end prototype

Versioned source and a working path through the interfaces needed for the experiment.

B

Reproducible test package

The package includes test configurations, fixtures, baselines, scoring code, data provenance, results, and documented validity limits.

C

Integration constraints

Documented interfaces, cybersecurity requirements, data rights, deployment limits, integration owners, and support assumptions.

D

Technical decision record

Results against thresholds, observed failures, unresolved questions, and the basis for the next phase.

CONTACT

Working on an SBIR topic with a difficult feasibility question?

Send us the topic, operational use case, key technical uncertainty, and due date. We can help define a practical prototype and test scope.

[email protected]