PreAccident Investigation Podcast

Host Todd chats with Mousa Yassin about shifting safety culture from blaming individuals to designing systems that tolerate failure and recover quickly. They cover life-saving rules, the concept of recoverability, lessons from software engineering like chaos testing, and the importance of learning over punishment.

The episode emphasizes practical ways to build resilient systems, nurture learning teams, and make safety training engaging and effective.

Todd Conklin tells the origin story of "learning teams," sparked by a self-reported near-miss at Los Alamos involving a postdoc and an arcing wrench. Rather than pursuing a punitive investigation, a group of workers gathered to identify what needed to be learned, uncovering broader gaps in postdoc training and safety planning.

The episode explains how learning teams prioritize asking better questions, collecting the right data, and designing system-focused solutions. Conklin describes how this approach spread across the lab and why it remains a fast, effective tool for operational improvement.

Todd Conklin talks with Corey Pitzer about fatality prevention, Human and Organizational Performance (HOP), and how safety thinking has shifted globally.

They explore controversial views—treating workers as problem-solvers, tensions between engineering/energy-based approaches and systemic/new-view thinking—and use real examples to show why designing systems that absorb variation matters more than trying to eliminate risk.

Todd interviews Professor David Woods about recent NASA mishaps and a growing cultural shift toward "cheaper, faster" decision-making that sacrifices safety. They explore how past safety gains have lost vitality, highlight cascading modern risks (the "messy nine"), and argue for mutual assistance and revitalized resilience practices.

Wood's most recent writing on this is available in the Bulletin of Atomic Scientists called:  Cheaper, Faster, and Who Gives a Damn about Anything Else.

The episode connects space, aviation, cloud outages, and AI-driven engineering to show why coordinated foresight and cross-disciplinary cooperation are essential to prevent far-reaching harm in today’s complex systems.

In this episode, Todd Conklin joins Amir Shahzad to discuss human and organizational performance, resilience, and how to design systems that allow failures to be caught and recovered before they become disasters. They explore the gap between work as imagined and work as done, the value of learning from everyday work, and practical steps leaders can take to create safer, more resilient workplaces.

They also cover cultural change, the role of procedures, adaptive behavior, and the potential—and risks—of AI in safety, all delivered with a mix of practical advice and light-hearted rapid-fire questions.