Por qué la mayoría de las recomendaciones nunca se implementan completamente.
Por Bruno Hounkpati
5 themes · 2 audiences · 4 languages — every article reads a real 2024–2026 incident.
Por qué los investigadores llegan a conclusiones antes de examinar la evidencia.
Cuando los procedimientos permanecen sin cambios después de las fatalidades.
El objetivo final: transformar la investigación reactiva en arquitectura de prevención.
The Bayesian and the limits of 'approved'
The Bayesian, read from the boardroom
BASF Geismar, and the authority that dissolved in a handoff
BASF Geismar, and the stop-work your board only thinks it has
Olin Freeport, PBF Martinez, and isolation that wasn't where the work was
Olin, PBF Martinez, and what your control-of-work record actually proves
Northrop Grumman, argon, and a confined space reclassified out of its controls
Northrop Grumman, a reclassified confined space, and the controls a board never sees removed
Moss Landing, thermal runaway, and a technology deployed faster than its playbook
Moss Landing, and governing a technology deployed faster than its failure could be contained
Análisis Tripod Beta impulsado por IA