3 ms·
It seems pretty easy to verify the blocks are stacked properly. First, make them out of a shape that naturally fits together. Then, build them with a vertical
by adrianmonk 8y ago
It seems pretty easy to verify the blocks are stacked properly.
First, make them out of a shape that naturally fits together.
Then, build them with a vertical hole (shaft) through them. Put a reflector under the bottom one. Before the crane releases a block, it shines a light down the hole and checks if it is reflected back. Misalignment will block the light. (Or, you can do something similar with electricity, connected contacts on the top and bottom, and a check that current flows through the entire stack.)
- kardos 8y agoSounds good except what do you do about biology that decides to take up residence in the holes?
- adrianmonk 8y agoI suppose you'd have to clean it regularly. When the crane goes to pick up a block, it could do the optical continuity check then too. If it fails, that indicates that either the block was moved after it was stacked correctly (unlikely but possible) or it got dirty. In that case, something comes along and cleans it. Maybe a rod-shaped brush (like a vacuum beater) that spins. Or a water jet. Also, you can have more than one hole for redundancy. If one of them is clear, that should be good enough. Hmm, or rather than holes through the middle, make them notches or grooves on the side and shine a laser through them.
- progfix 8y agoThat only works in theory. In practice the reflector, light, whatever gets dirty and you would have false alarms non-stop.
- saulrh 8y agoSlap a tiny solar-powered GPS unit and transmitter on each block. ~$10 of electronics is marginal compared to the $2000 of concrete for each block. If any break, you already have a crane that you can use to move the block out of the way for repair; just hand it over to an operator who can do the special-case operation by hand. If you get a beacon unit with a MTBF of 10 years and not awful clustering you might get one to five failing per stack per day, so you'd probably be able to handle a hundred stacks with a single operator dedicated to remote failure-handling.