An article at NewScientist caught my eye the other day and got me thinking. The title of the article was Blurry X-rays become clear with new software. 

I wonder if it would be possible to improve the current efficiency of today's instrumentation by scrapping the faster and more powerful is better route and moving towards a more energy efficient design.  This new design concept would be based on taking a measurement/scan/image as quickly and efficiently as possible and then adding a software step at a later stage in the experiment work flow to improve the result.

This direction might be a hard sell at the trade-shows but seeing how every other market is heading towards more energy efficient designs the instrumentation field should start planning now.

Oh, and if any large instrumentation companies with R&D money to burn would like to give this a go, we'd be more than happy to take this problem on.