PlantaOS Space Intelligence domain knowledge. Use for any task involving Freedom Index, building zones, sensor data, anomaly detection, digital twin, dashboard pages, simulation scenarios, or HORSE/Renault context.
F = P / D P (Perception) = log2(N) x T where N = active_sensors x avg_quality, T = depth D (Distortion) = 0.30xR + 0.10xO + 0.20xTb + 0.20xC + 0.20xM Boundaries: P >= 1.0, D >= 0.01 Classification: GREEN (F>2.0), YELLOW (1.0<F<=2.0), RED (F<=1.0) Gradient Law: dx/dt = -P(x) * grad(D(x)) — optimal flow direction
Freedom (configs) -> Logic (rules) -> Relations (correlations) -> Physics (sensors) Each anomaly is tagged with which layer causes it.
collective_F = sum(F[z]occ[z]) / sum(occ[z]) flow_efficiency = actual_gradient_flow / total_flow (0-1) bottleneck = zone where F < 1.0 on high-traffic path health = 0.40norm(collective_F) + 0.30flow_eff + 0.30(1-bottleneck_ratio) -> 0-100
18 zones, 2 floors, ~500m2 total See refs/building.md for complete zone specs and adjacency
Environment (temp, humidity, CO2, lux) | Resources (energy, water) Occupancy (presence, count, flow) | Equipment (HVAC, lighting) Safety (fire, doors) | Digital (network) | Reference (HSE bands) Events (weather, shifts) | Action (alerts, tickets) See refs/kpis.md for ranges and frequencies
Energy savings: -18% vs baseline Maintenance reduction: -27% via predictive alerts Space efficiency: +9% via occupancy optimization
occupancy <-> CO2 (r > 0.6) occupancy <-> energy consumption external_temp <-> HVAC power adjacent zones <-> temperature bleed shift_schedule <-> occupancy patterns HVAC_state <-> temperature <-> comfort <-> energy
"AFI", "Architecture of Freedom Intelligence", "FLRP" (use "layer analysis") "Thesis 1/2/3/4/5" (use the actual metric names)