A multi-year, multi-site apiary research programme quickly turns into dozens of overlapping trials — different queens, different treatments, different seasons — unless every experiment gets a stable ID and a locked-down plan before data collection starts. This scene renders that registry as a physical ledger: rows of floating record tokens, one column per site, one row per registered experiment, each stamped with a status the moment its preregistration is locked.
A registry entry doesn't have to forbid deviation — it just has to record the original plan immutably, so any later change is visible as a documented deviation rather than a silently rewritten hypothesis. That distinction is what makes a result credible.
A 3D ledger of apiary experiments across sites and seasons, where a sweeping timeline plane locks each entry's plan the moment it's preregistered, and reveals whether the finished trial matched that plan or deviated from it.
Each apiary site is a vertical rail of record tokens in registration order. The sweeping "now" plane locks tokens as it passes them, then colours each one green (confirmed as planned) or amber (flagged deviation) once results come in.
Set the number of sites, scrub the registry timeline to watch entries lock and resolve, adjust the deviation rate, and filter the ledger to preregistered-only or flagged-deviation entries to audit either group.
A registry doesn't need to prevent deviation from a plan — it just needs to timestamp the original plan immutably, so any later change is a visible, documented deviation rather than a silently rewritten hypothesis.