Audit is not logging
Why traceability fails even in systems with extensive logs and why auditability requires explicit decision context.
TrailMQ Wiki
Concepts, architecture notes and explainers around audit-first MQTT, regulated data flows, evidence trails and industrial messaging.
Browse by topic
All articles
Foundations
GMP is shaped by time
Why auditability in regulated systems is fundamentally a time-based architecture problem, not a documentation issue.
6 min read
Implications
Manufacturing is not behind in AI – it is behind in trust
Why AI adoption in manufacturing fails not because of models, but because trust, traceability and validated data foundations are missing.
6 min read
Foundations
Can an MQTT broker be GxP compliant?
Why MQTT itself is not GxP compliant or non-compliant and why compliance depends on auditability, data integrity and decision context around the br...
7 min read
Distinctions
Why context must exist at decision time
Why reconstructing context after the fact is not sufficient and how missing context undermines auditability and trust.
6 min read
Models
GMP is not binary – it scales with proximity
Why GMP requirements increase gradually with proximity to the physical process and how architecture often hides this gradient.
7 min read
Foundations
What we can learn from Pharma
Why Pharma's emphasis on traceability, ownership and explainability is not a burden but a blueprint for scalable automation and AI.
6 min read
Implications
AI is not the solution when the process is not understood
Why weak process understanding undermines AI effectiveness and why contextual clarity outperforms model complexity.
7 min read
Models
Decision-centric architecture
Why modern systems must treat decisions as first-class architectural elements instead of implicit side effects of data processing.
7 min read
Foundations
We can't automate trust
Why trust cannot be added after automation and must be embedded in industrial communication and data pipelines.
7 min read
Implications
Why AI does not break GMP
Why AI is not incompatible with GMP and how unclear architectural boundaries are the real source of compliance failures.
7 min read
Distinctions
Intervals are not abstractions – they are commitments
Why time intervals are not a technical detail but a formal commitment that defines meaning, responsibility and comparability.
6 min read
Models
The hardest problem in Industrial AI
Why context alignment across time, batches and process phases is harder than prediction and determines AI success.
7 min read
Implications
Why dashboards do not create trust
Why visibility without traceability fails to build trust in regulated and automated industrial systems.
6 min read
Foundations
What makes IIoT systems AI ready
Why AI readiness in IIoT is determined by validation, context handling and traceable data foundations, not by model connectivity.
6 min read
Distinctions
Logs are not audit trails
Why log streams fail compliance and what true audit trails must guarantee in regulated industrial systems.
6 min read
Models
Operational twins are about translation, not visualization
Why operational twins focus on translating physical process reality into comparable, auditable representations.
7 min read
Implications
Why data pipelines decide whether regulated AI will succeed
Why AI success in regulated manufacturing depends on data pipelines that preserve integrity, context and traceability end to end.
7 min read
Foundations
How to start with GMP in IIoT projects
A practical way to approach GMP in IIoT by tracing one data path and making trust explicit from the start.
6 min read
Implications
AI speed versus regulatory trust
Why accelerating AI without transparency creates risk and why continuous compliance must evolve alongside continuous learning.
6 min read
Foundations
IIoT GMP basics – why data flow alone is not enough
Why IIoT systems in regulated environments need traceability, validation and lifecycle control beyond basic connectivity.
7 min read