Beyond Crude Clock Rules: Liljegren WBGT Modeling and Deterministic Constraints in Indian Construction — Rishi Singh

The Failure of Conventional Clock Rules

Across Northern and Central India (Delhi NCR, Rajasthan, Uttar Pradesh), summertime temperatures frequently exceed 45°C. The conventional regulatory response is a blunt tool: a mandated site shutdown between **12:00 and 15:00**.

Our historical meteorological analysis using Open-Meteo ERA5 reanalysis data revealed two alarming empirical realities:

1. **Morning Thermal Strain**: Due to high overnight humidity and rapid solar angle escalation, hazardous Wet-Bulb Globe Temperatures ($> 30^\circ\text{C}$) are routinely reached by **10:30 AM**—exposing workers to dangerous heat strain during hours legally considered "safe." 2. **Indiscriminate Inefficiency**: Fully shaded indoor activities (e.g., electrical conduits, basement plumbing, interior drywall) that could proceed safely in controlled conditions between 12:00 and 15:00 are shuttered, while heavy outdoor rebar bending resumes at 15:15 PM when solar heat load remains severe.

Physical WBGT Calculation: The Liljegren (2008) Formulation

Wet-Bulb Globe Temperature (WBGT) is the international gold standard for assessing environmental heat stress on the human body. Unlike simple "Heat Index" (which assumes stagnant indoor air and shade), WBGT accounts for four physical factors:

- **Dry-bulb temperature ($T_a$)**: Ambient air temperature - **Natural wet-bulb temperature ($T_{nw}$)**: Evaporative cooling capacity of the human skin - **Black globe temperature ($T_g$)**: Radiant solar heat absorbed by solid matter - **Wind speed ($u$)**: Convective cooling across the body surface

For outdoor environments with direct solar irradiance:

$$\text{WBGT}_{\text{outdoor}} = 0.7 \cdot T_{nw} + 0.2 \cdot T_g + 0.1 \cdot T_a$$

For shaded or indoor environments with zero solar load:

$$\text{WBGT}_{\text{shade}} = 0.7 \cdot T_{nw} + 0.3 \cdot T_g$$

python

def calculate_wbgt_workability(
    dry_bulb_c: float,
    relative_humidity: float,
    wind_speed_ms: float,
    solar_radiation_wm2: float,
    metabolic_rate_watts: int
) -> dict:
    """
    Computes physical WBGT using the Liljegren formulation and maps
    to OSHA / NIOSH continuous work/rest cycles.
    """
    # 1. Natural wet-bulb temperature via iterative psychrometric balance
    t_nw = compute_natural_wet_bulb(dry_bulb_c, relative_humidity, wind_speed_ms, solar_radiation_wm2)
    
    # 2. Black globe temperature via radiative equilibrium
    t_g = compute_black_globe(dry_bulb_c, wind_speed_ms, solar_radiation_wm2)
    
    # 3. Outdoor composite WBGT
    wbgt_outdoor = 0.7 * t_nw + 0.2 * t_g + 0.1 * dry_bulb_c
    
    # 4. In-shade composite WBGT (solar radiation = 0)
    t_g_shade = compute_black_globe(dry_bulb_c, wind_speed_ms, solar_radiation_wm2=0.0)
    wbgt_shade = 0.7 * t_nw + 0.3 * t_g_shade
    
    # 5. Screen against OSHA continuous work limit for metabolic workload
    # e.g., Heavy rebar bending = 400W (threshold: 26.0°C WBGT)
    threshold = get_screening_threshold(metabolic_rate_watts)
    is_safe_continuous = wbgt_outdoor <= threshold
    
    return {
        "wbgt_outdoor": round(wbgt_outdoor, 1),
        "wbgt_shade": round(wbgt_shade, 1),
        "safe_continuous": is_safe_continuous,
        "rest_cycle_recommended": get_rest_cycle(wbgt_outdoor, threshold),
    }

Architectural Principle: Deterministic Safety Decides, AI Explains

On a construction site, worker safety cannot depend on the probabilistic whims of an LLM. If an AI hallucinates that "it is safe to pour concrete on the roof at 42°C because the air feels breezy," workers face heatstroke.

In **WorkWise**, we designed a strict architectural boundary:

[ Weather Sensors / Forecast ] ──► [ Deterministic Python Safety Engine ]
                                                 │
                                 ├── Liljegren WBGT Calculation
                                 ├── CAQM GRAP Regulatory Gate
                                 └── Crew Capacity DAG Scheduler
                                                 │
                               ┌─────────────────┘
                               ▼
               [ Ground-Truth Operational Plan ]
                               │
               ┌───────────────┴───────────────┐
               ▼                               ▼
       [ UI Heatmap Matrix ]         [ Amazon Bedrock LLM ]
                                               │
                                 (Translates ground-truth plan 
                                  into Hindi / English audio briefing)
                                               │
                                               ▼
                                 [ Guardrail Validator ]
                                 (Rejects claims promising '100% safety' 
                                  or overriding deterministic gates)

Conclusion

Software engineering in physical industries requires grounding algorithmic design in verifiable physics and statutory reality. By coupling rigorous deterministic models with intuitive field interfaces, we can transform environmental data into life-saving operational decisions.