Robbing a bank with a parachuted pony

10 February 2021

“Say you wanted to rob a bank. If you were serious about it, you’d figure out a plan that would actually get you in and out of the bank, with the money. It would probably not involve a pixilated parachuted pony. But if you were the Joker, you’d just be getting started, because with him, well … he really wants to steal that money, but he also wants to dazzle the world […].”
Steve Englehart [source]

This is not about the Joker’s desire to ignore risk. This is about the average engineer designing e. g. a robust pixilated, parachutable pony for a specific job. These (sometimes) dazzling products are subject to multiple uncertainties across their lifecycle. In order to cater for this randomness, the engineer includes certain factors – or margins – into the modelling of the product:

  • Safety margins are typically included in the requirements against the uncertainties of use.
  • Design margins will be built in against the uncertainties of design, such as the usual RFC’s and more.

For a successful design you can say, the Design Factor must always be equal to or exceed the required Factor of Safety and the Margin of Safety is greater than zero.
Hence, the Design Factors are to be calculated within the realm of the engineer. Some of the appropriate factors of safety are based on considerations like the predictability of the loads, the strengths, and the wear estimates. As well, the direct and indirect consequences of a product failure due to erroneous engineering will need to be accounted for. Hence, these factors see a range between safety factors of four or up to ten when sever financial impacts or potential danger for human life are implicit in the operation. The materials going into the design pose another relevant factor of randomness and may see generally higher values for brittle materials and – for example – lower ones for commonly used durable metals with ample fatigue data sets. The dilemma is becoming even more vivid when our parachuting pony gets airborn since the domain of aerospace engineering generally targets lower design factors simply because of the costs associated with structural weight.

Lastly, if we look at your pony at the end of the design process, it may include considerable overdesign (weight, volume, cost, manufacturing requirements etc.) which will impact both the initial production costs as well as later cost in operation.

On the other side, if we are too optimistic though as to the safety margins, our little Pony may go ballistic. And the Joker will not be amused…!

Yet, no need to turn rogue and rob your bank without the pony – just yet. If you come to realize that you may have too inefficiently applied your safety margins, there is always the option to more efficiently incorporate your risk, for example, by leveraging next-generation stochastic optimisation algorithms.