Evaluating the level of technical risk associated with an engineering project and putting in place a risk management strategy is crucial to the success of any engineering project. Anything that has not been done before will have some technical risk associated with it. The work done to address technical risks is called Research and Development (R&D). This is where a bunch of engineers set up experiments and do the research to see if the have-not-been-done-before things can be done. This should be done first, not last, in a project unless you like to spend money for the sake of it. A short story to illustrate the point:
Back in the 80's the holy grail of design was Top Down design. This involved breaking down the system design task from the top down into a number of subsystems or modules (such as power supply, cathode ray tube, signal amplifier and housing for a TV) and then each of these into modules (such as transformer, rectifier and filter for the power supply) and so on, until a complete upside down tree structure was identified. Then the theory was that if you addressed the design of each part, starting at the top with the major modules you could not go wrong. Then some wag posted the upside down tree structure for an anti-gravity machine. Hidden away, at the lowest level, was a small module named vector reversal module. Think about how much of the design would be complete before the engineers finally turned their attention to this module... Needless to say, an appropriate risk assessment of this project would have immediately identified this module as one of very high risk. The risk reduction strategy would have then focused engineering attention on this module first rather than last.