About Our Research

At the Smart, Reliable, and Interpretable Systems (SRIS) Lab, we fuse logic, physics, and learning to build trustworthy, high-performance systems for intelligent manufacturing and critical equipment.
Our lab advances a unified methodology—Signal Temporal Logic (STL), causal structure, and physics-guided priors—to achieve interpretable diagnosis, safe autonomy, verifiable machine learning, and fleet-level health management. We aim for exportable industrial software stacks, standards-ready audit trails, and measurable impact on reliability, yield, and safety.
Verifiable
Industrial-Grade
Interpretable

Neuro-Symbolic Foundation Models

Trustworthy Manufacturing AI

Multimodal industrial pretraining with STL-regularized objectives and causal priors for machine-checkable rationales and controller-aware outputs.

Causal KGs Rule distillation Cross-site transfer Uncertainty

STL-Driven Safe Autonomy

Dual-Arm & Mobile Manipulation

Plans and controllers synthesized under temporal-logic contracts. Reference governors guarantee safety; residual RL optimizes within verified tubes.

Contract mining Compositional proofs Vision-force skills Runtime monitors

Causal-Logic PHM for Critical Equipment

Energy • Transport • Aerospace • Petrochemical

Physics-guided features + causal graphs + STL abstractions to reconstruct propagation chains and recommend cost-aware interventions across fleets.

Order/TP features PHM ontology Causal invariants Fleet policies

Research Directions — Details

Click a section to expand. Each direction includes a lab-website–friendly introduction, theory-forward core idea, key tasks, representative pilots, metrics, and alignment with national strategy.

Introduction

We build the next generation of manufacturing AI by unifying data-driven learning with logic, physics, and causality. By embedding Signal Temporal Logic (STL) constraints and causal structure into the representation space, our models produce machine-checkable rationales and calibrated uncertainty, enabling cross-factory transfer and controller-aware recommendations.

Core idea (theory-forward)

Pretrain on multimodal industrial data (time series, events, logs) with STL-regularized objectives, causal priors from knowledge graphs, and physics-consistent invariants. Dual formulations turn logic satisfaction into differentiable losses; distilled symbolic rules align human audit with closed-loop control.

Key tasks

  • Cross-site adaptation via knowledge-based federated learning and domain-invariant causal anchors.
  • Rule distillation from embeddings to minimal STL clauses with margin guarantees.
  • Controller-aware heads that output safe set-points and alarms aligned with contracts.
  • Uncertainty quantification (e.g., conformal prediction) tied to logic satisfaction margins.

Pilots & Metrics

Flexible job-shops Robotic assembly Rotating machinery Rationale fidelity Zero/low-shot transfer Causal precision/recall

National fit: 工业软件自主可控 / 新质生产力。

Introduction

We engineer robot autonomy that is safe by construction and fast by design. Safety is specified up-front as temporal logic contracts and embedded throughout planning, learning, and control. The result is dual-arm and mobile manipulation that adapts to changing workcells while certifying task timing, force limits, and human–robot interaction rules.

Core idea (theory-forward)

A hierarchical stack where high-level plans and low-level controllers are synthesized and trained under STL specifications. Reference governors project actions onto contract-satisfying manifolds; residual RL optimizes performance inside verified tubes. Runtime monitors provide certificate-based intervention and graceful recovery.

Key tasks

  • Contract mining from expert procedures; automatic refinement to STL/contract forms.
  • Certifiable motion primitives with compositional proofs of safety and timing.
  • Vision–force multimodal policies trained with curriculum schedules respecting contracts.
  • Online contract monitoring with recovery policies and formal violation explanations.

Pilots & Metrics

Force-controlled assembly AGV–arm handoff Rapid changeover STL satisfaction Verified accuracy Near-miss reduction

National fit: 先进制造装备与智能机器人。

Introduction

We move beyond alarms to actionable causality in Prognostics & Health Management. By coupling physics-guided signal processing with causal graphs and temporal logic abstraction, we reconstruct fault propagation chains across components, systems, and fleets—turning heterogeneous data into interventions that save uptime, energy, and cost.

Core idea (theory-forward)

Fuse order/transfer-path–aware features with causal discovery and do-calculus–informed reasoning, then elevate events to STL templates that encode mechanism-consistent signatures. Perform inference on a multi-layer PHM knowledge graph to generate interventions (derating, maintenance windows) with transparent justifications.

Key tasks

  • Unified ontology linking events → mechanisms → actions across fleets and vendors.
  • Robust feature extraction under variable regimes; virtual decoupling of mixed excitations.
  • Domain generalization with causal invariants for cross-asset transfer.
  • Fleet-level decision support: counterfactual simulation and cost-aware maintenance policies with logic guarantees.

Pilots & Metrics

Wind & gas turbines Rail traction CNC & spindles Lead time↑ Forced outage↓ RUL precision↑

National fit: 能源安全 / 交通强国 / 高端装备可靠性工程。

Fundings