README.md

---
license: cc-by-4.0
task_categories:
  - tabular-classification
  - tabular-regression
  - time-series-forecasting
  - reinforcement-learning
language:
  - en
tags:
  - synthetic
  - industrial-real-estate
  - logistics
  - warehouse
  - supply-chain
  - operations
  - decision-support
  - anomaly-detection
  - forecasting
  - policy-optimization
  - capacity-planning
  - cold-chain
  - last-mile
pretty_name: Forge Industrial Intelligence Pack
size_categories:
  - 10K<n<100K
configs:
  - config_name: default
    data_files:
      - split: train
        path: forge_industrial_sample.parquet
---

# Forge Industrial Intelligence Pack (Sample)

**A synthetic industrial real estate, logistics, and operations decision-telemetry dataset for anomaly detection, forecasting, decision-support, and policy-optimization research.** Each row is a complete operational scenario lifecycle — from signal emergence through telemetry evolution, detection, forecast, impact, and recommended action — with 14 top-level context fields tying market, tenant, facility, portfolio, and decision-owner layers together.

Built by [SolsticeAI](https://www.solsticestudio.ai/datasets) as a free sample of a larger commercial pack. 100% synthetic. No real facility data, no real tenant names, no real portfolio telemetry — market names (Dallas-Fort Worth, LA, Chicago, etc.) refer to generic submarket archetypes, not specific properties or operators.

## What is included

| File | Rows | Format | Purpose |
|---|---:|---|---|
| `forge_industrial_sample.parquet` | 10,000 | Parquet | Columnar, typed, best for analytics |
| `forge_industrial_sample.jsonl` | 10,000 | JSON Lines | Streaming / LLM training friendly |

**This sample:** 10,000 operational scenario lifecycles.  
**Severity tiers:** `medium`, `high`, `critical` (~3,300 each)  
**Status:** `decision_pending`, `confirmed`, `triaged` (~3,300 each)  
**Scenario classes (7):** `port_spillover_surge`, `power_constrained_ev_yard`, `labor_gap_service_degradation`, `cold_chain_excursion_risk`, `tenant_hypergrowth_overrun`, `last_mile_cutoff_compression`, `site_selection_power_arbitrage`  
**Markets (6):** Dallas-Fort Worth, Los Angeles, Chicago, Atlanta, New Jersey, Phoenix  
**Tenant industries (5):** ecommerce, 3PL, retail distribution, cold chain, industrial manufacturing  
**Action library:** 14+ playbook actions across labor, power, capex, tenant, network, and pricing levers

## Record structure

Each record is one operational scenario lifecycle with 14 top-level fields:

| Field | Type | Contents |
|---|---|---|
| `schema_version` | string | Pack schema version (`1.0.0-forge-industrial-sample`) |
| `event` | struct | `id`, `trace_id`, `timestamp`, `scenario`, `severity`, `status`, `confidence` |
| `organization` | struct | `sector`, `market`, `submarket`, `region`, `environment`, port/airport proximity, grid headroom, vacancy, base rent, development cycle |
| `identity_context` | struct | `principal_id`, `decision_lineage[]` (role/action/authority), `auth_method`, `stakeholder_latency_hours`, `meeting_load` |
| `vulnerability` | struct | `class`, `risk_taxonomy[]`, `exposure`, `severity_model` (base_score + vector) |
| `tenant_context` | struct | `tenant_id`, `industry`, `truck_profile`, growth rate, square footage, expansion openness, retention priority, churn risk |
| `facility_context` | struct | Building characteristics, capacity, energy profile |
| `portfolio_context` | struct | Network-level dynamics, capacity distribution |
| `telemetry_stream` | list<struct> | Ordered telemetry readings (dock utilization, queue length, power load, service commitment risk, etc.) with per-step event labels |
| `detection` | struct | `analytic_family`, `primary_risk_class`, `rule_logic`, `baseline_deviation`, `anomaly_score`, `confidence_band`, `signal_conflicts[]` |
| `forecast` | struct | Forward-looking predictions (demand shift, occupancy, pricing) |
| `impact` | struct | Economic consequences (revenue delta, NOI, capex, churn risk) |
| `response` | struct | `recommended_actions[]`, `primary_action`, `primary_action_score`, `primary_action_reason`, `alternative_actions[]`, `decision_owner`, `execution_window_days`, `playbook_id`, `capex_gate_required`, `stakeholders[]`, `expected_operational_outcome`, `recommended_tradeoff`, `execution_risk_band` |
| `simulation` | struct | `synthetic`, `engine`, `causal_coherence`, `friction_profile`, `ground_truth_label`, `intended_use[]` |

See [SCHEMA.md](./SCHEMA.md) for the full nested field breakdown.

## Why this dataset is useful

Most public operations / logistics datasets are either flat sensor-stream snapshots or narrow single-scenario studies. Industrial real estate telemetry (dock utilization, grid headroom, tenant churn signals, decision lineage) is effectively never published. This pack is shaped around what operational-intelligence and decision-policy teams actually need:

- Full scenario lifecycles rather than isolated sensor frames
- Balanced severity and status distributions across decision-pending, triaged, and confirmed cases
- Coupled detection → forecast → impact → response layers per record (not just telemetry)
- Economic impact fields (revenue, NOI, capex, churn) for decision-aware modeling
- Decision-lineage metadata for policy-interpretability research
- Cross-scenario coverage (port, power, labor, cold chain, tenant, last-mile, site selection)
- Compact parquet (~3.8 MB) that fits inside CI and notebooks, with a fuller JSONL for LLM training

## Typical use cases

- Operational-anomaly detection models
- Capacity-planning and forecasting pipelines
- Decision-support and recommendation engines
- RL / policy optimization for industrial operations
- Tenant-churn and retention modeling
- Site-selection and network-optimization research
- LLM fine-tuning on operational narratives and decision rationale
- Scenario-mining for logistics debugging
- Executive-dashboard and BI template development

## Quick start

```python
import pandas as pd
import pyarrow.parquet as pq

df = pq.read_table("forge_industrial_sample.parquet").to_pandas()

# Severity distribution
print(df["event"].apply(lambda e: e["severity"]).value_counts())

# Scenario × primary action cross-tab
df["scenario"] = df["event"].apply(lambda e: e["scenario"])
df["primary_action"] = df["response"].apply(lambda r: r["primary_action"])
print(pd.crosstab(df["scenario"], df["primary_action"]))

# Average impact.revenue_delta_usd by market
df["market"] = df["organization"].apply(lambda o: o["market"])
df["rev_delta"] = df["impact"].apply(lambda i: i.get("revenue_delta_usd"))
print(df.groupby("market")["rev_delta"].mean().round(0))
```

Streaming form:

```python
import json

with open("forge_industrial_sample.jsonl") as f:
    for line in f:
        lifecycle = json.loads(line)
        # one operational scenario lifecycle per line
```

## Responsible use

This dataset is intended for **research, model development, and decision-policy simulation** around industrial real estate, logistics, and operations. It contains synthesized telemetry, decision lineages, and economic-impact rollups — it does **not** contain real facility telemetry, real tenant identities, real portfolio NOI, or identifiable stakeholder data. Models trained on this data should be validated against real operational data under appropriate organizational review before being used for live recommendations or capex decisions.

## License

Released under **CC BY 4.0**. Use freely for research, operations-tool prototyping, education, and commercial development with attribution.

## Get the full pack

This Hugging Face repo is a **10K-lifecycle sample**. The production pack scales to 5M+ lifecycles with expanded market and submarket coverage, richer per-step telemetry (ms-level sensor traces where applicable), additional scenario classes (ESG signals, geopolitical supply disruption, labor automation events, bonded warehouse flows), multi-year longitudinal traces per tenant, parquet + JSONL + BI-tool-aligned delivery, and buyer-specific variants.

**Self-serve (Stripe checkout):**
- [**Sample Scale tier — $5,000**](https://buy.stripe.com/7sY5kD2j85QTfSb5lfeEo03) — ~25K records, one subject, 72-hour delivery.

**Full pack + enterprise scope:**
- [www.solsticestudio.ai/datasets](https://www.solsticestudio.ai/datasets) — per-SKU pricing across Starter / Professional / Enterprise tiers, plus commercial licensing, custom generation, and buyer-specific variants.

**Procurement catalog:**
- [SolsticeAI Data Storefront](https://solsticeai.mydatastorefront.com) — available via Datarade / Monda.

## Citation

```bibtex
@dataset{solstice_forge_industrial_pack_2026,
  title        = {Forge Industrial Intelligence Pack (Sample)},
  author       = {SolsticeAI},
  year         = {2026},
  publisher    = {Hugging Face},
  url          = {https://huggingface.co/datasets/solsticestudioai/forge-industrial-pack}
}
```