timebased statistical chain wear detection

my solution is mostly procedural, as opposed to adding expensive equipment. however, there is an essential cleverly designed simple piece of hardware that gets added to the top of the drivetrain (more about that later). 1. each chain link must be numbered sequentially for indexing 2. chains should be wiped free of debris whenever possible 3. my test procedure must be run before and after each workday my chainwear detection method is statistical and based on offline automated computerized comparison of the data acquired daily by human employees using the following procedure. 1. clean the chain thoroughly 2. make sure the 5 sensors on my special cog are clean. the part is hereafter refered to as the "pentacog" 3. load the crane with the known test load (a lead weight at maximum capacity for the crane) 5. cycle the chain so that link#1 is on the pentacog 6. turn on the recorder unit, followed by the hoist at lowest speed 7. run until end of the chain is reached 8. the recorder data is then uploaded into an ongoing database. what happens next is completely automated and is based on heuristic analysis of the time between sensor blips. what the test procedure records is the amount of time it takes for each link segment to cross it. elongation of chains will cause small fluctuations in the beat. rather than inspecting individual links, my system is more interested in partitioning the chain into probable areas of defect. this can yield recommendations from the automated system for a group of chains to undergo more thorough human inspection. because data is compared over the short term and longterm, trends can mathematically indicate specific wear signatures. from seemingly simple data, profound predictions can be made. the longer the data is collected, the bigger the database, and the more accurate predictions will be. the pentacog is nothing more than a toothed gear of special geometry that aligns each chain and optically detects when 1 chain has spun through. it is engage