344 single-point failures

I highly recommend watching the superb documentary Unknown: Cosmic Time Machine to increase your appreciation of human achievement. 🛰️

It follows a team of engineers and scientists on an ambitious mission to launch the James Webb Space Telescope (JWST) and the near impossibility of the project.

How many single-point failures do you have around you?

Landing on Mars: 90 – one of the hardest things to do
James Web Telescope: 344 – the largest of any space mission (225 in the sun shield deployment)

JWST design started in 1996, construction in 2004, testing in 2017 and final launch in 2021 for an estimated cost of US$10bn. Initial estimates were for launch in 2007 for a cost of US$500m.

Post one test, NASA discovered critical fasteners did not work; the fix took six months and cost US$150m.

The impossible was possible, and JWST produced the most detailed and deepest infrared images of the universe, capturing galaxies that formed just a few hundred million years after the Big Bang. 🌌

Enjoy the images here:

🏗️ Back on Earth, we are surrounded by large, complex projects; almost all will be over budget and over time. To put some data around that assertion, read How Big Things Get Done: The Surprising Factors Behind Every Project’s Success (and Failure) by Bent Flyvbjerg, dubbed “the world’s leading megaproject expert” and Dan Gardner.

The book dives into the challenges and lessons of delivering large-scale projects, from bridges and skyscrapers to tech rollouts and film productions. It explores why so many grand initiatives end up delayed, over budget, or failing altogether while also detailing what separates successful ones.

In his dataset of 25 project types covering 16,000 projects, he details mean cost overruns and mean cost overrun if the project is in the tail of more significant overruns:

Nuclear Storage: 258% – 427%
Olympic Games: 157% – 200%
IT: 75% – 447%
Buildings: 62% – 206%
Rail: 39% – 116%
Wind Power: 13% – 97%
Solar: 1% – 50%

Good news for renewable energy, less so for Olympic host cities.

The book argues that projects designed in modules—independent, manageable segments—are more likely to succeed. Modularity allows project parts to be completed in phases, tested, and adjusted without jeopardizing the entire project. Projects that lack this flexibility face higher risks when encountering unforeseen issues.

🫵 The final words in the book: Know that your biggest risk is YOU and your behavioural biases and shallow thinking.

😡 Recently, my fence contractor installed four new panels, three identical and the other not. Logic would suggest that all would be the same. Logic and competence aside, I could have included the words “identical panels” in the scope of work and been onsite to monitor progress.

My work mantra has always been “only expect what you inspect”…so the failure could be mine. 🤦‍♂️