Planning for complex systems lacks the convergence metrics and defect detection that manufacturing achieved through six decades of statistical process control. FDRP addresses this gap by treating each planning decision as a manufactured artifact with measurable convergence and gate-reviewed quality, applying SPC, 5S, RAMS, Andon, and configuration freeze directly to planning. Applied across 62 production runs (1,477 decisions) with 32 active subsystems, the system's strongest empirical contribution is the antimatter building programme: 68 LLM-generated expert specialists across two rounds produced 28,936 lines of analysis and 1,656 structured findings, discovering expert persistence as an emergent operational pattern. Cross-model verification with 3 independent LLMs revealed 4 systematic bias patterns, suggesting that N≥3 models are needed for blind spot detection. The architecture has since grown to 557 database tables, 188 views, 80 system rules, and 645 knowledge index entries. We propose progressive disclosure as the unifying architecture—a hypothesis under test.
FDRP Research Paper
Download
Download Paper (PDF, v21.0)
SHA-256:
dd7fab470f1c7c9804d49d64700844057c2fb7d15b54282c84c39f117d14b78f
Abstract
Table of Contents
Act I: The Framework
- Introduction: The Planning Quality Gap
- Related Work and Theoretical Foundations
- Architecture: FDRP as Manufacturing System
- The Convergence Velocity Tensor
- Progressive Disclosure as Unifying Principle
Act II: The Evidence
- Production Dataset: 62 Runs Across Multiple Domains
- Case Study: Antimatter Building Programme
- Case Study: Asymmetric Cyber Defence
- Case Study: Fire Crisis Response
- Cross-Model Verification
- Cross-Domain Validation
- Expert Expansion and Persistence
- Bias Detection and Mitigation
- Statistical Analysis and Control Charts
Act III: The Reflection
- Limitations and Threats to Validity
- The Self-Correction Thesis
- Future Directions
- Conclusion
Cite This Work
BibTeX
@article{olos2026fdrp,
title = {FDRP: A Self-Evolving Architecture for Planning Quality},
author = {Olos, L.},
year = {2026},
note = {Preprint, v21.0},
url = {https://fdrp.liviu.ai/paper/FDRP-Paper-v21.pdf}
}
APA (7th edition)
Olos, L. (2026). FDRP: A Self-Evolving Architecture for Planning Quality. Preprint, v21.0. https://fdrp.liviu.ai/paper/FDRP-Paper-v21.pdf
Chicago (17th edition)
Olos, L. “FDRP: A Self-Evolving Architecture for Planning Quality.” Preprint, v21.0, 2026. https://fdrp.liviu.ai/paper/FDRP-Paper-v21.pdf
Version Archive
| Version | Date | Key Changes | Download |
|---|---|---|---|
| v21.0 | Post-R1027 update: learning collector, data quality auditor, traceability metric scoped + backfill, daemon effectiveness refinements | Current | |
| v20.0 | 9 new subsystems (#24-#32): cognitive architecture, daemon taxonomy, thinking chains, wiring verification, limit hunting. 557 tables + 188 views, 32 subsystems total | ||
| v14.0 | Visualization pass: 4 new figures, 19 fixes, concept2image disclosure | ||
| v13.0 | Comprehensive 4,472-line version with full evidence | ||
| v8.0 | Wave 8: cross-model reviewed, 10 fixes | ||
| v6.0 | First complete draft with all sections |
Supplementary Materials
Full technical report, raw data exports, and CERN department one-pagers
are available in the gated section. Credentials: username
fdrp-reader, password published in the paper footnote.