According to an Earth First! Journal article, one of the ways to solve the current crisis of worldwide bee decline is just to move on and build new, improved bees. Rather than fix the old, boring, evolution-bred biotic bees, we can use human-made machines which emulate bees and some of their functions -- like pollination.
Pollination, performed flawlessly by bees for millennia, is in jeopardy due to declining bee populations, also known as colony collapse disorder (CCD). Although the cause(s) for CCD are not known, some of the most likely candidates are toxic pesticides and genetically-modified crops. In case you don't recognize the connection, modern agriculture (not to mention plant biodiversity) relies on the process of pollination to succeed. To put it more bluntly, without bees and other pollinators, humans can't grow food and we die.
Human-engineered agriculture (may have) caused this, so human engineering can get us out of this, goes the thought. I won't mention the massive biotech company named in the Earth First! article so I don't get sued, but I'm sure you've heard of them.
Although these robobees aren't swarming yet, this technology may soon be made possible thanks to research on Micro Air Vehicles at the Harvard Microbiotics Lab. The researchers believe that they may be able to autonomously pollinate fields of crops using these Robobees.
I love this project from the engineering side of it and I encourage you to read more about it (and don't get too hung up on the military applications just yet.)
Many of the advances in engineering and computer science are simply about humans learning to emulate the advanced, efficient processes nature has developed. This is a good thing. However, the use that is described (to pollinate GMO crops) has a great many drawbacks, some obvious and some not.
I think that it is selfish and naive to think that the problems that have been caused by the method of growing crops that biotech companies have developed will be solved so easily. It seems like such a simple, elegant solution, but as we've seen many times, the Earth's biosphere is a much more complex ecosystem than we can normally comprehend, much less duplicate perfectly. This applies even to relatively small ecosystems like cash crop agriculture.
Yes, it may solve one of the immediate problems (lack of bees to pollinate), but that problem indicates an inherent sickness in the agriculture system that is only partially "technical" in nature. It has more to do with greed and impatience than technical problems to overcome.
One other drawback, not always thought about when engineers are concerned, is that nature inherently has inefficient processes which have the potential to produce unplanned benefits. It is this inefficiency which human engineers never try to replicate (because it is the opposite of what they attempt to do), but that can provide both intangible and tangible unexpected benefits.
I'm sure that the designers of the Robobees didn't have merely a pollination replacement in mind when they set out to create this marvel of circuitry and software. (Yes, read further; the software to emulate and mimic the collective intelligence of bee colonies is as amazing as the microengineering.) I'm excited to see the other applications these devices could have to benefit all of us.
As for the bees, I hope that humans will put as much energy and thought into preserving this crucial and beautifully designed organism as we do designing its robotic doppelganger.