⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.
A 3D smart-hive monitor where a machine-learning anomaly model scores simulated weight, temperature and acoustic sensor drift against a colony's own baseline, raising an early swarm-risk flag - and showing exactly where that flag can be wrong.
Most practical hive AI is pattern-recognition on sensor time series, comparing a colony against its own established baseline rather than a universal "normal hive". Swarm preparation shows up as a falling weight trend, less stable temperature and a rising acoustic queen-piping score - combined here into a single z-score-based risk percentage.
Choose a scenario - Normal season, Pre-swarm buildup, or Noisy sensors - then adjust simulated time speed and detection sensitivity. Watch the risk score, the AI verdict and the scrolling sensor chart, and see how a more sensitive threshold catches real risk earlier but also produces more false alarms from ordinary noise.
An early-warning flag is valuable for prioritising which hives to inspect first, but it is not a substitute for physically checking for queen cells - real deployments still need a beekeeper's own judgement to confirm anything the model flags.