Predicting the Future is a Perilous Business

Predicting the future is a perilous business with many traps for the unwary. Are we asking the right questions of technology roadmapping teams?
— by Dr. Francis Mullany, Director of Roadmapping
Inside iNEMI's roadmaps over the last quarter century, one can see the mixed results of technology roadmapping. A simple case in point was highlighted in a recent study by Dartmouth. The projected timelines for minimum inter-pin pitches had a fairly consistent expected curve shape from 0.5 to 0.2mm over a 4- to 10-year timeframe … except the curve kept shifting outward in time, from roadmap to roadmap. The issue was not so much that we didn't expect the technology to get there, it's just that the conditions, either in terms of the R&D investment needed or market pull, didn't line up in the timeframe suggested.

And then, of course, things get even more, ahem, challenging, when you are looking to prophesy revolutionary change. The cellular industry certainly didn't forecast the huge surge in demand for 3G wireless data when the iPhone arrived on the scene in 2008. They had to scramble to ensure that their radio access network infrastructure wasn't overwhelmed with traffic loads increasing by 50% or more per year.

The interesting aspect of the iPhone story was that there was nothing hugely revolutionary about the hardware components. The main innovations were in the user experience and the app store model. The hardware technology was mature and ready to go – other factors held back the demand, factors such as software and business models.

Technology roadmaps have to do more than paint a picture of the destination – that's the role of vision papers.  Roadmaps have to outline how we are going to get there. But, what's the alternative to predicting a series of milestones with dates attached?  

As iNEMI launches the content creation cycle for its new online roadmap, we are exploring other approaches. While we will still attempt to predict timings, we are looking at alternative metrics for measuring technology progression. Two such metrics with a long track record are that of the TRL (technology readiness level), originally proposed by NASA and later adopted by the EU, and the MRL (manufacturing readiness level), proposed by the U.S. Department of Defense.  

For roadmapping, the great advantage of readiness levels is that once you know today's readiness level for a particular technology, you can identify what kind of R&D investments are needed to reach the next level. For example, going from MRL level 5 (“prototype components in a product relevant environment”) to level 6 (“prototype system”) requires clarity in the interfaces between components. Cross-industry standardization may be needed at this point, to enable the corresponding supplier ecosystem. More generally, an experienced R&D manager will even give you a good order of magnitude estimate for the time and resources needed to ensure the transition between levels. The other benefit is that TRLs and MRLs allow for a decoupling of technology-related predictions from market-related factors like end-user demand and business models.   

For the iNEMI Roadmap, we won't abandon attempts to put a timeline on technology advancements. However, we will use MRLs and TRLs as a complementary means of mapping out the future of key technologies. Our definition will nearly completely mirror the current U.S. definitions, with minor adaptations for electronics manufacturing. Having a standardized set of metrics for technology maturity will sit well with our new approach of building technology roadmaps as a structured database of drivers, needs and capabilities.

Hopefully, a new set of signposts will help us avoid the perilous pitfalls of technology forecasting.

Interested in being part of the cross-industry expert team for creating the next generation of the iNEMI Roadmap?  Please fill out the short expression-of-interest form!