Four glowing platforms are the areas a real environmental audit checks: the air emissions stack, the wastewater outfall, the solid-waste yard and the chemical storage compound. Each one has a fluctuating measured level and a regulatory limit; when the level rises above the limit, its ring turns red — a non-conformance. Small drones are the audit team: they circulate between the four areas, dwell to take a sample, and log a finding the moment they catch a station in breach. If auto-remediation is on, a logged finding triggers a corrective action that permanently lowers that station's baseline emissions — mirroring PDCA's "Act" step — so the compliance rate should climb back up over repeated audit cycles instead of staying flat.
level(t) = base·(1 − fixes·0.18) + drift + incident·spike(t)
violation ⇔ level(t) > limit·strictness
finding logged on dwell-start if violation is true
- Audit team size — how many auditors work the site in parallel; more auditors catch a transient breach sooner.
- Regulatory strictness — shifts every station's limit down (stricter, left) or up (more lenient, right); the same emissions profile can be compliant under one regime and a violation under another.
- Incident rate — how often a station gets a random spike (a spill, a malfunction, an unplanned discharge) on top of its steady baseline.
- Sampling pace — how quickly auditors move between stations and complete a dwell/sample cycle.
- Auto-remediate — toggles whether a logged finding is followed by a real fix (lower baseline) or just recorded with no correction, so violations keep recurring.
Real-world relevance: this is why ISO 14001 treats an audit as a cycle, not a one-off inspection — internal or external audits only have value if findings actually convert into corrective action; otherwise the same non-conformance shows up again at the next audit.