Landing Congestion Analysis

Explore the impact of landing shock events on stochastic playback outputs for the synthetic medium scenario.

[1]:
import sys
from pathlib import Path

PROJECT_ROOT = Path.cwd().resolve()
while PROJECT_ROOT != PROJECT_ROOT.parent and not (PROJECT_ROOT / "pyproject.toml").exists():
    PROJECT_ROOT = PROJECT_ROOT.parent
if not (PROJECT_ROOT / "pyproject.toml").exists():
    raise RuntimeError(
        "Notebook must be executed within a FHOPS checkout (pyproject.toml not found)."
    )
if str(PROJECT_ROOT) not in sys.path:
    sys.path.insert(0, str(PROJECT_ROOT))

from docs.examples.analytics import utils
from fhops.scenario.synthetic import SyntheticDatasetConfig, sampling_config_for

SCENARIO = PROJECT_ROOT / "examples/synthetic/medium/scenario.yaml"
ASSIGNMENTS = PROJECT_ROOT / "docs/examples/analytics/data/synthetic_medium_sa_assignments.csv"

config = SyntheticDatasetConfig(
    name="synthetic-medium",
    tier="medium",
    num_blocks=8,
    num_days=12,
    num_machines=4,
)

baseline_sampling = sampling_config_for(config).model_copy()
baseline_sampling.landing.enabled = False
baseline_tables, baseline_sampling = utils.run_stochastic_summary(
    SCENARIO, ASSIGNMENTS, sampling_config=baseline_sampling
)

shock_sampling = baseline_sampling.model_copy()
shock_sampling.landing.enabled = True
shock_sampling.landing.probability = 0.3
shock_sampling.landing.capacity_multiplier_range = (0.35, 0.7)
shock_sampling.landing.duration_days = 3
shock_tables, shock_sampling = utils.run_stochastic_summary(
    SCENARIO, ASSIGNMENTS, sampling_config=shock_sampling
)

Baseline Ensemble Snapshot

[2]:
baseline_tables.shift.head()
[2]:
day shift_id machine_id machine_role sample_id production_units total_hours idle_hours mobilisation_cost sequencing_violations blackout_conflicts available_hours utilisation_ratio downtime_hours downtime_events weather_severity_total
0 1 S1 M1 tethered_harvester 0 6.925670 5.989121 0.000000 0.0 0 0 0.0 NaN 2.010879 1 0.0
1 2 S1 M1 tethered_harvester 0 12.660000 8.000000 0.000000 0.0 0 0 8.0 1.000000 0.000000 0 0.0
2 2 S1 M4 loader 0 0.000000 3.712075 4.287925 0.0 0 0 8.0 0.464009 4.287925 1 0.0
3 4 S1 M2 tethered_shovel_or_skidder 0 7.605535 4.806026 3.193974 0.0 0 0 8.0 0.600753 3.193974 1 0.0
4 5 S1 M2 tethered_shovel_or_skidder 0 0.000000 4.079769 3.920231 0.0 0 0 8.0 0.509971 3.920231 1 0.1

Landing Shock Ensemble Snapshot

[3]:
shock_tables.shift.head()
[3]:
day shift_id machine_id machine_role sample_id production_units total_hours idle_hours mobilisation_cost sequencing_violations blackout_conflicts available_hours utilisation_ratio downtime_hours downtime_events weather_severity_total
0 1 S1 M1 tethered_harvester 0 4.006046 5.989121 0.000000 0.0 0 0 0.0 NaN 2.010879 1 0.0
1 2 S1 M1 tethered_harvester 0 7.322980 8.000000 0.000000 0.0 0 0 8.0 1.000000 0.000000 0 0.0
2 2 S1 M4 loader 0 0.000000 3.712075 4.287925 0.0 0 0 8.0 0.464009 4.287925 1 0.0
3 4 S1 M2 tethered_shovel_or_skidder 0 7.322980 4.806026 3.193974 0.0 1 0 8.0 0.600753 3.193974 1 0.0
4 5 S1 M2 tethered_shovel_or_skidder 0 0.000000 4.079769 3.920231 0.0 0 0 8.0 0.509971 3.920231 1 0.1

KPI Comparison

[4]:
import pandas as pd

comparison = pd.DataFrame(
    {
        "scenario": ["baseline", "landing_shock"],
        "mean_production": [
            baseline_tables.shift.groupby("sample_id")["production_units"].sum().mean(),
            shock_tables.shift.groupby("sample_id")["production_units"].sum().mean(),
        ],
        "mean_downtime_hours": [
            baseline_tables.shift["downtime_hours"].mean(),
            shock_tables.shift["downtime_hours"].mean(),
        ],
        "mean_utilisation": [
            baseline_tables.shift["utilisation_ratio"].mean(),
            shock_tables.shift["utilisation_ratio"].mean(),
        ],
    }
)
comparison
[4]:
scenario mean_production mean_downtime_hours mean_utilisation
0 baseline 33.138204 2.278553 0.705316
1 landing_shock 23.094934 2.278553 0.705316

Production Distributions

[5]:
baseline_prod = baseline_tables.shift.groupby("sample_id")["production_units"].sum()
baseline_chart = utils.plot_distribution(
    baseline_prod, title="Baseline Production Distribution", xlabel="Production Units"
)
baseline_chart
[5]:
[6]:
shock_prod = shock_tables.shift.groupby("sample_id")["production_units"].sum()
shock_chart = utils.plot_distribution(
    shock_prod, title="Landing Shock Production Distribution", xlabel="Production Units"
)
shock_chart
[6]: