Case study · Defence R&D
Measuring how ready a technology really is.
With a premier Indian Institute of Management, BLG built a fuzzy-logic assessment platform for a research organisation under the Ministry of Defence — replacing spreadsheet-driven Technology Readiness Level reviews with a structured, auditable, role-based system informing a project portfolio worth lakhs of crores.
- Client
- Defence R&D organisation · Ministry of Defence
- Academic partner
- A premier IIM
- Engagement
- Full-stack build
- Deployment
- Secure internal network
The challenge
Maturity was a matter of opinion.
Technology Readiness Level is the standard 1–9 scale research organisations use to track how close a technology is to deployment. Rigorous in theory — in practice, it was decided in a meeting.
Across labs, TRL scoring lived in disconnected spreadsheets and PDFs, filled in from memory rather than evidence. A “TRL 5” from one lab and a “TRL 5” from another weren’t measuring the same thing — and neither had a record of why.
The approach · score the ambiguity
Instead of forcing one discrete number, TRL is modelled as a fuzzy set — a technology can sit 92% in TRL 5 and 8% in TRL 4 at once. Eleven evaluation factors, anchored to organisation-defined targets, produce a normalised score passed through overlapping membership functions — a probability-weighted reading instead of a guess dressed up as a fact.
The system
Three roles, one accountable pipeline.
Lab Director
System-wide administrator
Owns the question bank, creates projects, assigns project directors, and holds visibility across every lab's TRL reports.
Project Director
Project-level orchestrator
Builds the work-breakdown structure (system → sub-system → component), assigns scientists, and schedules recurring assessment cycles.
Scientist
Assessment executor
Runs the guided assessment across 11 factors, attaches evidence per question, and submits a response set that locks permanently once scored.
Scientists walk the 11 factors one at a time — typed questions, conditional follow-ups, evidence uploads — with progress autosaved. Nothing is scored until final submission, at which point the assessment locks and becomes a permanent record.
What it automates
Removing the manual chase.
What used to be a periodic scramble before a review is now a continuous, self-scheduling process — nobody has to take a TRL number on faith.
- Recurring assessment cycles defined once — the scheduler generates the next cycle automatically
- Automatic alerts to scientists and directors 48 hours before a deadline
- Rollup reporting: project → schedule → technology → individual, always drillable
- TRL progression tracked over time, not as a single snapshot
- One-click export to PDF and Excel for offline review
- Every score traces to a specific answer, scientist, and piece of evidence
Engineering
Built for a security-conscious internal deployment.
- Authentication
- JWT with 8-hour expiry · bcrypt-hashed credentials
- Authorisation
- Role enforcement per route + data scoping by project ownership
- Transport
- HSTS, frame and content-type protections on every response
- Files
- Capped, authenticated uploads with path-traversal-safe handling
- Auditability
- Locked assessments are immutable; every scheduler run is logged
Stack
Runs on an internal network, not the open internet — so the architecture optimises for auditability, data integrity, and role isolation over public-scale traffic.
The outcome
From a spreadsheet opinion to a defensible score.
- ₹L Cr+
- Project portfolio informed by assessments
- 11
- TRL factors standardised org-wide
- 1–9
- Fuzzy-scored scale with probability weights
- 3
- Accountable role tiers, lab to bench
Every assessment produces the same structured evidence trail, every score is computed the same mathematical way — and every level of the organisation, from the scientist filling in the form to the lab director reading the rollup, is looking at the same numbers.