Many safety challenges are seemingly beyond the ability of current techniques and technologies to resolve, because technologies to resolve them have not yet been developed, are only now being conceived of, or are currently infinitely expensive. One important way to deal with these is to sneak up on them.

Sneaking up involves understanding the odds of disaster in detail (statistically), and driving them down to their lowest possible value given these three constraints. Mathematically there is still tiny risk that someone will be hurt, but practically — in the real world — it doesn’t happen because the circumstances for the disaster are so rare that they don’t come to pass over the time period being managed.

This only works if continuous improvement is foremost in the plans for the future. Successive design upgrades are required, each pushing safety a little further forward. Without this push, if progress stagnates, statistics will catch up and the disaster will eventually occur.

It’s very important to stay ahead of the curve, yet it can be done very successfully a little at a time.