In one such case now, I’m looking at biological switches as a source of inspiration. Cells and organisms often have a need to be in one of two states, and they need a switching function that is all-or-nothing and yet not overly sensitive to fluctuations in the switching signal. A “normal” reaction of the sort
Substrate + signaling agent --> Output
is difficult to control as it relies on first or second order kinetics, and you can't esablish a critical threshold for the reaction to initiate and stay there. Instead, a better choice is to set up two opposing reactions against each other.
A terrific example is sex determination (open access, but registration might be required). Humans and most other mammals, (platypuses seem to be an exception, which isn’t too surprising as they seem to be the exception to nearly every rule regarding mammals) start out with the potential to become either sex. As the embryo is growing, asexual gonads develop. At some critical point, if the Y gene is present, the SRY gene will encourage the development of the gonads into testes instead of ovaries. However, two other genes are critical in making the switch complete: WNT and FGF9, both of which work in opposite directions when the SRY is present or absent.
In cases where one or more of these switching genes is absent, then things can get muddled with the result being a mix of sexual traits. But when things work properly, the signal is processed extremely well and most people end up as clearly male or female. (I've greatly simplified all of this - The Scientist article has much more details on sexual determination, and the first two cited articles are great for probing mathematical models of this.)
Clearly a neat approach to solving chemical switching.