Reusable simulation architecture for dynamic relational learning environments

Design · All designs
Evidence · no claims cited

Description

The article abstracts SupplyNet's pipeline into "a reusable simulation architecture for dynamic relational learning environments." A graph represents domain entities, dependencies, and time-varying states; agents and learners observe relevant subgraphs, take domain-specific actions, and trigger state transitions; a database stores temporal snapshots supporting replay, branching, and comparison. The reusable layer separates from a domain-specific layer so other subjects can reuse the architecture by replacing the domain schema, transition rules, agent heuristics, event types, and learning metrics.

Design Implications

Context

Requirements

• Subjects must be modelable as dynamic relational systems involving multiple connected entities, local decisions or disturbances, delayed propagation of effects, and trade-offs that learners need to diagnose over time.

Constraints

• The architecture was designed and evaluated only in the supply chain management context; cross-subject instantiations are discussed as future work.

Target Learners

• university students in domains modelable as dynamic relational systems

Learning Goals

• diagnosing delayed propagation of effects and trade-offs in interconnected systems

Claims

Related Designs

• Supplynet Visual Simulation System
• Behavior-based persona and course-grounded memory architecture for simulated students
• CogEvolution framework: ICAP perception, IRT-based memory retrieval, and evolutionary cognitive state update
• Multi-agent AI tutoring architecture with an orchestrating Tutor Agent and specialist agents

Examples

-

Key Sources

• Yanjia Li, Kelcy Kexin Han, Tianrui Hu, Yi-Fan Cao, Huamin Qu, Sicheng Song. (2026). SupplyNet: Supporting Visual Exploratory Learning in Supply Chain via Contextual Multi-Agent Simulation. https://arxiv.org/abs/2606.24694

Page metadata

Field	Value
type	design
id	reusable-simulation-architecture-dynamic-relational-learning
title	Reusable simulation architecture for dynamic relational learning environments
description	The article abstracts SupplyNet's pipeline into "a reusable simulation architecture for dynamic relational learning environments." A graph represents domain entities, dependencies, and time-varying states; agents and...
status	review
generated	by process:settle-candidates on 2026-10-11
trust	unverified — no one has confirmed this page's content
sources	1 — yanjia-li-2026

Frontmatter is the page's machine-readable record — see the Schema & Guide.