Failure analysis for turbine blades. lithium batteries. sinkholes. pipeline ruptures. corrosion damage. semiconductors. bearings. composite delamination.

Where the world's most specialized knowledge of technical failures converges.

13+industries
80+research areas
300+experts
How it works

Expert answers — in a fraction of the time.

The Institute brings together scientific literature, engineering knowledge, leading experts, and AI-assisted research to help organizations understand complex failures and identify the expertise required to solve them.

Traditional
Weeks of referrals, RFPs & vetting
With FAI
Minutes to answers

Explore Knowledge

Consult the Institute's AI Research Concierge, trained on a curated body of technical literature, Institute publications, standards summaries, and reference materials. Start for free.

Consult Experts

Access specialists from academia, industry, government laboratories, consulting practice, and engineering leadership. Find a match.

Assemble Expertise

Some questions require multiple disciplines. The Institute can identify, recruit, coordinate, and manage expert teams tailored to your project. Tell us more.

get started

What failed?

Start a conversation with our AI Research Concierge trained on thousands of hours of privately held research authored by the world's top experts. Select a category to prompt the Research Concierge.

AI Research ConciergeMine our knowledge base or find an expert
Start by exploring topics to the left, or tell me what you'd like to investigate. I'll ask you a few questions so I can direct you to the best possible research or expertise.
Find an expert

Expertise Unlocked

The Institute can conduct structured AI-assisted interviews with multiple specialists across relevant disciplines, allowing you to compare perspectives before deciding how to proceed.

Who we serve

Our Industries

We serve board members, executives, attorneys, officials, and researchers across more than a dozen industries.

research report no. 1 · by russ rosenzweig, executive director

Known in Advance

A structural analysis of the one hundred worst engineering failures of the modern era, and the expertise that was never assembled.

99 of 100
The knowledge needed to prevent the disaster existed before it happened.
67 of 100
The fatal mechanism was described in writing, before the event, and overruled.
1 of 100
Genuinely beyond the knowledge of its time. Nobody knew.
read the report
news & insights

What's failing — and why it matters.

Institute analysts track high-profile failure events as they develop.

See all insights
common questions

What counsel ask before they trust a site like this.

These come before the engineering does. If the answers are unsatisfying, nothing further down the page will help.

What does the Institute actually do?
It puts the knowledge of why things fail in one place, and makes the first hour of it free. Describe a failure and the research concierge will help you understand what the likely mechanisms are, which engineering disciplines investigate them, what evidence matters and what is at risk of being lost. Where a matter needs a retained expert, the Institute finds the right one.
Will you tell me what caused my failure?
No. Nothing here, including concierge conversations, is an opinion on cause, defect or liability. Naming the mechanisms consistent with a description is not the same as concluding which one occurred — that requires examination of the actual evidence by a qualified engineer, and it is the difference between orientation and an opinion.
How is the Institute paid?
The research assistance is free and stays free. When a matter requires a retained expert — for an examination, an advisory engagement or testimony — the Institute arranges the engagement through its expert network and is compensated for that work. That is the entire business model, and we publish it because visitors who cross-examine people for a living should not have to guess.
The evidence is about to be moved or repaired. Does that matter?
Enormously, and it is the most urgent thing on this page. Physical evidence is perishable in a way documents are not: fracture surfaces corrode, components get cleaned, scenes are cleared and equipment is repaired or scrapped in the ordinary course. Once altered, the analysis that would have been possible is often simply gone. If anything is scheduled to be moved, repaired or discarded, raise it immediately.
How often was a failure genuinely unforeseeable?
Rarely, on the Institute’s own research. A structural analysis of the hundred worst engineering failures of the modern era found that in 99 of 100, the knowledge needed to prevent the disaster already existed before it happened; in 67, the fatal mechanism had been described in writing beforehand and overruled. One was genuinely beyond the knowledge of its time. The recurring failure is not of engineering knowledge but of assembling it.
Which side do you work for?
Both, matter by matter. The Institute serves plaintiff and defense counsel, insurers and insureds, manufacturers and owners. Physics does not change according to who asked, and an institution that only ever reached conclusions convenient to one side would not be worth consulting from either.

Something failed. Find out why.

Tell us what happened. We'll triage it and connect you with the right expert — usually within one business day.

talk to an expert
AI Research ConciergeMine our knowledge base or find an expert
Tell me what failed and I'll help scope it — the likely mechanisms, what's at stake, and which expert fits. I won't give the formal analysis here.