OntoL: Ontology-Driven Reasoning Platform for Autonomous Blue Force Simulation
OntoL is an ontology-based intelligent reasoning platform that integrates with general simulation systems to enable autonomous blue force decision-making through three modules: OntoL-Data for multi-source intelligence fusion, OntoL-Manage for simulation orchestration, and OntoL-Bank for case knowledge reuse, demonstrated via a 15-step anti-submarine warfare simulation loop.
Overview
OntoL is an ontology-based intelligent reasoning platform designed to connect with general simulation platforms, providing full-process wargaming capabilities for intelligent blue forces across multiple combat scenarios. The platform comprises three core modules that work in concert:
OntoL-Data : Multi-source heterogeneous intelligence data system that ingests, normalizes, and fuses intelligence, hydrological, equipment, geographic, and open-source data into a semantic knowledge network capable of reasoning.
OntoL-Manage : Simulation sandbox engine responsible for scenario editing, force management, intelligent agent scheduling, situation driving, adversarial simulation, firepower closure, and visualization dashboards.
OntoL-Bank : Knowledge asset library that archives historical simulation cases, intelligence evidence chains, tactical solutions, adversarial samples, and effectiveness metrics to support similar case retrieval, tactical review, and experience reuse.
After integration with a general simulation platform, red and blue intelligent agents no longer follow fixed scripts but can observe, understand, reason, decide, and dynamically adjust like real commanders. The entire combat confrontation loop—from scenario construction and route planning to intelligence fusion, combat plan generation, detection and localization, firepower strike, and effectiveness evaluation—runs end-to-end on a unified foundation.
A Complete Simulation: 15-Step Closed Loop (Collaborative Version)
Using a typical scenario—"Three red submarines of a certain type depart from a port, transit the Bashi Channel, and proceed to a target sea area to execute a mission"—OntoL connects to the general simulation platform and decomposes the complete intelligent anti-submarine confrontation into an interlocking simulation chain. The three modules participate throughout.
Phase 1: Scenario Construction, Planning Behind the Scenes (OntoL-Manage Lead)
Step 1: Scenario Editing & Route Planning – In the visual sandbox editor, routes, depths, and speeds are planned for submarines, patrol aircraft, UAVs, and nuclear attack submarines; red and blue forces are managed on a single screen.
Step 2: Simulation Launch – Timeline-driven, supporting multi-speed execution and playback; all-platform force situations are driven uniformly by OntoL-Manage and dispatched to the connected general simulation platform for entity motion calculation.
Phase 2: Ontology-Based Intelligence Analysis, Making the System "Understand" (OntoL-Data Core Capability)
This is OntoL's core perception and judgment phase. The multi-source heterogeneous intelligence data system automatically aggregates and aligns three categories of evidence:
Step 3: Sonar contacts (raw underwater acoustic anomaly signals ingested)
Step 4: Satellite imagery time-series data
Step 5: Home port berth time-series comparison, identifying submarine departure behavior
Step 6: Open-source video, social media, and other OSINT cross-verification
Multi-source intelligence feeds into the OntoL-Data ontology intelligence system reasoning network for automatic fusion, outputting enemy situation judgments, trajectory predictions with uncertainty cones, and key threat assessments. All conclusions carry traceable evidence chains; data normalization, semantic alignment, and conflict verification are handled by the multi-source heterogeneous intelligence data system. Judgment results are pushed in real time to OntoL-Manage sandbox for situation updates and simultaneously dispatched to the general simulation platform.
Phase 3: Intelligent Plan Generation, Automatic Task Allocation (OntoL-Manage Decision Scheduling)
Based on the situation judgments from the multi-source heterogeneous intelligence data system, OntoL-Manage automatically generates anti-submarine patrol plans: UAV forward warning, patrol aircraft zonal search, with task airspace, ingress/egress routes, and coordination sequencing generated in one click.
Step 7: Generate patrol plan
Step 8: Dispatch plan to general simulation platform for patrol execution; intelligence to combat action requires only one confirmation.
Phase 4: Multi-Means Collaborative Detection, Deep Sea "Perspective" (OntoL-Data Perception + OntoL-Manage Entity Driving + General Simulation Platform Calculation)
Step 9: Temperature-measuring buoys first probe seawater temperature profiles, real-time calculation of sound velocity distribution, providing ocean environmental basis for sonar selection and deployment (environmental data ingested into multi-source heterogeneous intelligence data system).
Step 10: Passive sonar buoys (VLAD arrays) conduct wide-area listening to lock suspected bearings.
Step 11: Active sonar buoys precisely localize and converge.
Step 12: Magnetic anomaly detection finally confirms underwater target and maintains continuous tracking.
From "suspected contact" to "target confirmation," the detection system progresses layer by layer; various detection observation data are returned by the general simulation platform and continuously written into the multi-source heterogeneous intelligence data system to refresh situation awareness.
Phase 5: Threat Ranking, Precision Strike, and Case Archiving (OntoL-Manage Strike Closure + OntoL-Bank Archiving)
Step 13: Target threat ranking – system comprehensively ranks red platform threat levels based on target position, speed, depth changes, and intent confidence, making command focus immediately clear.
Step 14: Generate strike plan – match weapons status for nuclear attack submarine, recommend strike trajectory and cooperative allocation.
Step 15: Dispatch strike command to general simulation platform to complete torpedo strike: after plan confirmation, firepower closure; full-process events, evidence, and metrics fully recorded.
After the simulation round ends, OntoL-Bank automatically archives the entire simulation asset : raw intelligence, situation time-series, decision chains, detection records, strike logs, and effectiveness metrics, forming a searchable adversarial case library that supports tactical review, rapid reuse of similar scenarios, and model iterative training.
OntoL Core Capabilities
OntoL-Data Multi-source Heterogeneous Intelligence Data System: Ontology-Driven Multi-source Intelligence Fusion Engine Intelligence is no longer isolated messages and measurement points but a structured, reasonable semantic knowledge network. Completes normalization, correlation, and conflict verification of acoustic signatures, satellite, hydrological, and open-source intelligence. Judgment conclusions carry complete evidence chains; reasoning process is credible and traceable. Supports multiple data source access.
OntoL-Manage: Full-Chain Sandbox Simulation and Intelligent Agent Collaboration Engine Connects to general simulation systems for scenario editing, force sandbox, situation driving, plan generation, detection interaction, firepower strike, and real-time visualization. Outputs decision commands to the general simulation platform for entity dynamics and environmental model calculation; blue and red intelligent agents autonomously perceive, maneuver, collaborate, and dynamically adjust actions. Platform supports rapid extension, adapting to anti-submarine, air defense, shore-sea confrontation, and other combat scenarios.
OntoL-Bank: Combat Case and Tactics Asset Library Automatically precipitates complete cases, intelligence samples, tactical solutions, and effectiveness evaluation metrics from each simulation. Supports case retrieval, tactical comparison, equipment argumentation traceability, and training review, converting simulation adversarial experience into reusable knowledge assets.
Realistic Battlefield Fog-of-War Reasoning Target trajectory predictions include uncertainty metrics; intelligence exhibits fuzziness, conflict, and latency characteristics, restoring cognitive limitations under real reconnaissance environments and avoiding idealized judgments.
Scientific Quantified Effectiveness Evaluation System Each simulation automatically outputs a standardized indicator set, supporting tactical doctrine verification, new equipment argumentation, and command personnel adversarial training evaluation.
Signed-in readers can open the original source through BestHub's protected redirect.
This article has been distilled and summarized from source material, then republished for learning and reference. If you believe it infringes your rights, please contactand we will review it promptly.
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.
