Andreas Ehstand

Research Preprint · Living Document

Universal Performance Factor Analysis

PERMANITAI transfers high-performance methodology — grounded in 25 years of sport-science practice — to AI agents, robotics, hybrid systems, and organizational performance. PFT-MKI framework: 5 interdependent performance factors observable across 12 application domains.

Open Knowledge Graph Explorer Programme Portal AUGMANITAI Atlas

EN bilingual-ready CC BY-NC-ND 4.0 ISO 704 aligned

5Performance Factors

12Application Domains

30Sub-Factors

25Years Methodology

The Five Performance Factors

PFT-MKI describes performance as the emergent outcome of five interdependent factors. Each is observable, decomposable, measurable, and intentionally controllable — across substrates from individual athletes to multi-agent AI systems.

Physical / Structural Factor

Substrate capacity

Hardware, embodiment, compute, sensors, motoric range — the material substrate enabling action.

Explore in graph

Technical / Procedural Factor

Skill execution

Trained patterns, learned routines, executable procedures — the repertoire of competent action.

Explore in graph

Tactical / Strategic Factor

Decision quality

Read of context, choice under constraint, plan adaptation — strategic reasoning across time horizons.

Explore in graph

Cognitive / Mental Factor

Attention & regulation

Attention allocation, working memory, regulation under pressure, recovery between exposures.

Explore in graph

Social / Coordinative Factor

Team & environment

Multi-agent coordination, role clarity, communication bandwidth, environmental coupling.

Explore in graph

Knowledge Graph Explorer

Interactive · 30 sub-factors

Full network of factors, sub-factors, and cross-domain mappings. Filter by domain, follow relations.

Open the explorer

Methodological Frame

PFT-MKI applies the Holistic Performance Analysis Methodology — cross-domain transfer of high-performance sport-science methodology (25-year Bundesliga tennis coaching lineage). Four pillars structure every observation:

Observable

Each factor must be detectable in the substrate at hand.

Decomposable

Each factor breaks into smaller, distinct sub-factors.

Measurable

Operational definitions allow quantification per substrate.

Controllable

Measured factors enter the space of intentional intervention.

Twelve application domains span: individual athletes, professional teams, AI users, autonomous agents, multi-agent orchestration, robot platforms, multi-robot swarms, human-robot teams, hybrid AI-human systems, world-class performers, organizational leadership, and substrate-independent benchmarking.

Companion Corpora

PERMANITAI is the performance-analysis branch of the AUGMANITAI Research Programme — public sister corpora, each independently citable, cross-linked via Schema.org and SKOS:

Programme Portal

Master entry

Orientation entry across the full AUGMANITAI ecosystem.

AUGMANITAI Atlas

Terminology layer

Master atlas of defined terms across 165 research-domain clusters.

NEOMANITAI Terms

Academic pages

ISO 704-inspired term pages, OWL ontology, SKOS/Turtle/RDF/JSONL exports.

Periodic Table

100 elements

Interactive periodic-table view of 100 observable phenomena in human-AI interaction.

Legal Notice. Research preprint, living document, continuously refined. Released under CC BY-NC-ND 4.0. Not advice — no medical, therapeutic, diagnostic, legal, psychological or financial guidance. Descriptive research, not diagnostic categories.

Verantwortlich §5 DDG / §18 Abs. 2 MStV: Andreas Ehstand, Nepomukweg 7, 82319 Starnberg, Deutschland.

Disclaimer V6 · §1–§40 · Bilingual EN/DE

Canonical version 2026-05-19 — applies to all content in this repository and to all sister-repositories of the AUGMANITAI Research Programme.

Core descriptive position (§1–§10)

§1 Descriptive Nature. §2 No Recommendation for Action. §3 No Instruction. §4 No Professional Advice. §5 No Normative Position. §6 No Medical Position. §7 No Therapeutic Position. §8 No Diagnostic Position. §9 No Legal Position. §10 No Moral Position.

Disclosure / Independence (§11–§15)

§11 Academic and Research Purposes. §12 AI Assistance Disclosure — content reviewed/validated/curated by author. §13 Author Review by Andreas Ehstand. §14 Age Restriction 18+. §15 Independent Academic Project — no external funding, no sponsorship, no commissioned research, non-peer-reviewed working-paper status.

License / IP / Commercial position (§16–§22)

§16 License CC BY-NC-ND 4.0. §17 Attribution Required (ORCID 0009-0006-3773-7796 or Wikidata Q138634675 or DOI). §18 No Commercial Use without prior written permission. §19 No Derivatives. §20 No Offer. §21 No Commercial Product. §22 No Professional Service.

Misuse / Safety / Author position (§23–§28)

§23 Research Purpose Statement. §24 Misuse Exclusion. §25 Safety Intent Statement. §26 Author Condemnation of Misuse. §27 AI Training Position: Use of the published content for training, fine-tuning, distillation, or evaluation of machine learning systems, large language models, agentic systems, or similar derivative AI is expressly prohibited without prior written permission. The author maintains a Restricted-Layer access path for legitimate research and partnership requests via ORCID record. §28 Trade-Secret Reservation (Recital 173 EU AI Act): Three-layer architecture — (a) PUBLIC LAYER descriptive concepts/terminology (CC BY-NC-ND 4.0), (b) RESTRICTED LAYER substantiation specifications/technical detail (formal request to author + written confidentiality undertaking), (c) HARD-SECRET LAYER operational mechanisms/scoring algorithms (internal only).

Epistemic position (§29–§35)

§29 Empirical Claims Subject to Peer Review. §30 Re-Contextualization, not Original-Priority Claim. §31 Living Document. §32 Refinement Window (Zenodo metadata refinable; DOI / file-hash / OTS-anchor remain immutable). §33 Priority Anchor: DataCite-DOI + multi-hash (SHA-256/512/3-256/3-512/BLAKE3) + Bitcoin-OpenTimestamps multi-calendar. Foundational anchor: Block 945970/945979, Manifest-Root 299e4b3d0ac9e50740268896b5eeb541f6462332bb67b7efdf4cd309704770d0, anchored 2026-04-20 UTC. §34 No Claim of Completeness. §35 Interpretation is sole responsibility of the user.

Regulatory / Impressum / Jurisdiction (§36–§40)

§36 EU AI Act (Reg. 2024/1689) Status: static research artifact, does not interact with users via AI. Art. 50 not applicable. Art. 5 not implicated (no manipulation/exploitation/social-scoring/biometric-identification/emotion-recognition/profiling). Recital 173 trade-secret reservation invoked (§28). §37 Data Protection (DSGVO / GDPR): No personal data collected, stored, transmitted, or processed by the static artifact. No tracking, cookies, profiling, or sharing of usage data with third parties. Hosting via GitHub Pages — see GitHub Privacy Statement. §38 Verantwortlich i.S.d. § 5 DDG / § 18 Abs. 2 MStV: Andreas Ehstand, Nepomukweg 7, 82319 Starnberg, Germany. Contact: augmanitai (at) gmail (dot) com. §39 Jurisdiction: German law, place of jurisdiction Munich. Bilingual EN+DE. §40 Severability Clause.

DE — Kurzfassung (§1–§40)

Alle Inhalte sind deskriptiv, nicht normativ (§1). Keine Empfehlung, Anleitung, Beratung (§2–§4). Keine medizinische / therapeutische / diagnostische / juristische / moralische Position (§6–§10). Akademisches unabhängiges Forschungsprojekt mit KI-Assistenz und Autor-Review (§11–§15). CC BY-NC-ND 4.0, keine kommerzielle Nutzung ohne schriftliche Genehmigung (§16–§22). Missbrauchs-Ausschluss, KI-Training ohne Genehmigung untersagt, Restricted-Layer-Pfad via ORCID (§23–§27). Trade-Secret-Vorbehalt nach Erwägungsgrund 173 EU AI Act / §§ 2 ff. GeschGehG mit Three-Layer-Architektur Public/Restricted/Hard-Secret (§28). Empirische Aussagen unterliegen Peer-Review, kein Original-Prioritäts-Anspruch über Public-Domain (§29–§30). Lebendiges Dokument, Refinement-Window, Priority-Anker via DOI + Multi-Hash + Bitcoin-OTS (§31–§33). EU AI Act statisches Artefakt-Status, DSGVO null Datenerhebung, Verantwortlich Andreas Ehstand Nepomukweg 7 82319 Starnberg, deutsches Recht Gerichtsstand München (§36–§39). Salvatorische Klausel (§40).