Estimating R&D schedule and cost: all the reasons not to, and one reason to do it anyway

Summary form only given as follows. R&D cost money, a great deal of money; yet the spenders of that money are rarely asked to commit to a cost-to-complete statement or to a time-to-finish pronouncement. And when asked, they object on grounds that R&D is essentially a journey into the unknown-off limits to cost estimation and schedule commitment. This assertion should be refuted on grounds that a credible R&D estimate may always precede the actual completion of the project, thus offering some decision freedom to the budgeting authority. R&D practitioners are fond of claiming that estimates inhibit the innovative spirit. This argument overshadows the notion that estimates would allow a better use of scarce R&D funds, thus offering a chance to young innovators and others who trail the allocation priority list. Another argument against estimating the effort to achieve an R&D objective is based on the premise that no one except the leading researcher/developer can accurately render an estimate. This can be seriously challenged by the methodologies developed in the last decade by AI practitioners, and knowledge engineers interviewing human experts for developing expert systems. In short: R&D cost estimation and schedule appraisal can be done credibly, and doing so will draw more resources to this activity, which all agree is the corner stone for economic prosperity