alt Hacker NewsI agree that it adds complexity, but it seems to me that the minimum there is on the order of the joints on a lawn chair, a screw drive, and a stepper. Not exactly high tech and if sufficiently overspec'd not likely to fail even in extremely adverse conditions.
The thing is that most solar projects are commodity energy projects optimized for cost per watt, not necessarily reliability or efficiency (watts per area). We're not building them for the constraints of Mars bases or fancy spacecraft, just run of the mill generation.
Thus the tracker has to not just pay for itself, but be a better investment than simply buying more modules. The goal isn't reliability at any price, but cost effectiveness compared to making the project a little bigger and adding a few more modules.
Even with home rooftops (which have limited space), it's much cheaper to buy more premium modules with higher efficiency, or optimize their outputs with per module micro inverters or optimizers, than to mount the array on a tracker. With big commercial installations, there's even less of a space premium.
Trackers aren't just a one time upfront cost of manufacturing either, but an ongoing maintenance cost over the lifetime of the project. They get more expensive as less and less of the industry uses them and fewer parts and labor are available for them. They just can't compete with the Chinese dumping of cheap modules, which are so cheap now that almost all non Chinese manufacturers have gone bankrupt. So the industry is left with a glut of modules, which are probably just going to get cheaper and more efficient, while tracking technology and costs haven't improved anywhere near the same amount. It's all relative, and the trackers are losing big time.
Sure, as a science project with infinite funding, maybe you can make a super reliable one. In the real world, though, installations across the world have found them largely unappealing from a cost effectiveness standpoint.