Digital Twin Real-Time Process Simulation 2026
Digital Twin Real-Time Process Simulation 2026
Real-time digital twin simulation is moving from conceptual frameworks to operational deployment, with 60+ retrieved records spanning 2019–2026 capturing convergence of IoT, edge compute, and AI-driven model calibration. Model synchronization has emerged as the defining engineering challenge across the dataset.
Bidirectional Physical-Virtual Coupling at Operational Speed
Digital twin real-time process simulation (DT-RTPS) is defined across the retrieved dataset as the bidirectional, data-driven coupling between a physical system and its virtual counterpart, where the virtual model continuously ingests live sensor data, executes simulations, and optionally feeds control decisions back to the physical layer. The dataset spans 60+ records from 2019 to 2026.
Five core sub-domains are identifiable within this dataset: discrete-event simulation-based digital twins for manufacturing and logistics; physics-based and CFD real-time simulation using reduced-order models; data-model-driven hybrid simulation; co-simulation architectures for virtual commissioning; and predictive clone twin sequencing that runs temporally offset copies to forecast future system states.
Key enabling technologies identified across the dataset include IoT and IIoT sensor infrastructure, cloud and edge computing, AI and ML for model calibration, FMU (Functional Mock-up Unit) interfaces, LiDAR and SLAM for spatial reconstruction, and large language models for model generation. These enablers are converging to make sub-second physical-to-virtual data loops commercially viable.
Innovation in this dataset is concentrated among a handful of large industrial players rather than distributed across a long tail of smaller assignees. Siemens (across Aktiengesellschaft, Corporation, and Ltd. China entities) accounts for 5 filing entries in retrieved records, followed by Fujitsu Limited and ETRI each with 4 filings in this dataset, consistent with the capital intensity of real-time industrial simulation systems.
Filing Trends and Technology Cluster Distribution
The retrieved dataset shows three identifiable phases of activity from 2019 to 2026: a foundational phase (2019–2020) dominated by literature and architecture frameworks, a development and scaling phase (2021–2023) with the largest patent filing cluster, and a maturation phase (2024–2026) marked by AI-native and federation-oriented filings.
Patent Filings by Technology Cluster — Dataset Snapshot
Automated synchronization and closed-loop IoT simulation account for the most distinct patent entries in this dataset, each with 4 or more retrievable filings, reflecting their centrality to commercial deployment readiness.
↗ Click bars to exploreFiling Activity by Phase (2019–2026) — Dataset Snapshot
The development and scaling phase (2021–2023) contains the largest cluster of patent filings in this dataset, with the maturation phase (2024–2026) introducing AI-native and federation-oriented patents as new directional signals.
↗ Click bars to exploreKey Deployment Domains for Digital Twin Real-Time Simulation
The retrieved dataset spans six identifiable application domains, with manufacturing historically dominant but recent 2024–2026 filings demonstrating rapid diversification into telecommunications, financial IT infrastructure, cleanroom environments, and live entertainment.
Manufacturing and Industrial Production
The dominant application domain in this dataset, with records spanning production line commissioning, virtual factory environments, and reconfigurable manufacturing systems. The 2019 real-time co-simulation platform for virtual commissioning established the pattern of running real control technology against simulation models before physical deployment. A 2025 Unreal Engine-based manufacturing monitoring patent from India signals game-engine infrastructure entering industrial real-time simulation.
Industrial SimulationInfrastructure and Built Environment
China University of Mining and Technology-Beijing filed two US patents in 2025 for a utility tunnel digital twin applying real-time CFD-based simulation with a reduced-order model and calibration algorithm. The Locus Cell Co. cleanroom monitoring digital twin (US, 2026) applies real-time simulation to contamination risk prediction in controlled manufacturing environments. An intelligent campus system (literature, 2022) demonstrates further extension to civic infrastructure.
Infrastructure MonitoringTelecommunications and 6G Networks
BTS Corporate Information Technologies filed a WO patent in 2024 claiming a digital twin-based system for 6G EDuRLLC (event-defined ultra-reliable low latency communication) services. Multiple literature records in the dataset identify 6G network management as a nascent but significant DT application domain. This domain is characterized by the need for ultra-low latency simulation of network states, distinguishing it from industrial use cases.
Network SimulationFinancial Services IT Infrastructure
Bank of America Corporation filed a US patent in 2024 for a “digital twinning data simulator” that applies real-time digital twin simulation to IT infrastructure redundancy: when a primary computer system goes offline, the digital twin generates simulated outputs based on last-known state data to maintain service continuity. This represents one of the first financial-sector entries in the retrieved dataset. Meta Live Inc. also filed patents in WO (2024) and IN (2026) for real-time digital twins of live events in VR environments.
IT ResilienceKey Patent Assignees in Digital Twin Real-Time Simulation (Retrieved Records)
In this dataset, Siemens entities collectively account for 5 filing entries across WO and US jurisdictions — the highest count in retrieved records — spanning closed-loop IoT simulation, semantic modeling, and automated ML-guided synchronization. Fujitsu Limited accounts for 4 filing entries in this dataset, concentrated in a tightly related patent family targeting the predictive clone twin sequencing problem across US and EP jurisdictions.
Top Assignees by Filing Count in Retrieved Records (Dataset Snapshot)
↗ Click bars to exploreSiemens (Aktiengesellschaft / Corporation / Ltd. China)
Siemens entities account for 5 filing entries in retrieved records, spanning WO and US jurisdictions across filings from 2022 to 2026. Technology areas include closed-loop IoT simulation for industrial plant assets (Siemens Aktiengesellschaft, WO 2022 and US 2024), digital twin modeling and simulation methods (Siemens Ltd. China, US 2022), and automated ML-guided visual synchronization that detects spatio-temporal differences between as-built and as-planned models (Siemens Corporation, WO 2024 and US 2026). The 2026 US filing represents the most recent patent in the dataset.
Germany / United StatesFujitsu Limited
Fujitsu Limited accounts for 4 filing entries in this dataset, with filings from 2022 to 2024 across US and EP jurisdictions. All filings relate to the predictive clone twin sequencing approach: creating temporally offset clone digital twins driven by a source twin’s output via a data stream synthesizer node to forecast future system states in IoT networks. This tightly related patent family is the only multi-jurisdiction, multi-year filing family in the dataset targeting the temporal forecasting problem specifically. Status includes granted and pending records.
Japan — US, EPFive Directional Signals from 2024–2026 Filings
The most recent filings in the dataset (2024–2026) reveal a shift from architecture-definition patents toward AI-native twin generation, automated visual synchronization, cloud elasticity for simulation compute, federation for multi-system interoperability, and spatial AR/VR integration with real-time simulation.
LLM and Generative AI-Driven Twin Generation
Chaos Industries (US, 2025) claims a system using generative transformer networks and large language models to generate digital twins with realistic simulation of real-world conditions. This represents a fundamental shift: rather than manually engineering simulation models, LLMs are used to synthesize model structure from available data, potentially reducing deployment time and cost significantly. IP strategists should monitor this space for blocking positions as generative model-to-twin pipelines mature.
Automated ML-Guided Visual Synchronization
Siemens Corporation’s 2026 US filing and 2024 WO filing describe a system that uses ML models to ingest raw visual data, generate an as-built twin, compare it against the as-planned twin, and automatically update the planned model to reflect real-world deviations. This eliminates the need for manual model maintenance in large-scale infrastructure contexts. The 2026 filing is the most recent patent entry in the retrieved dataset, signaling that spatio-temporal difference detection is becoming a core engineering concern.
Closed-Loop IoT Simulation vs. Predictive Clone Twin Sequencing
Click any row to explore further.
| Dimension | Closed-Loop IoT Simulation (Siemens) | Predictive Clone Twin Sequencing (Fujitsu) |
|---|---|---|
| Primary Assignee | Siemens Aktiengesellschaft / Corporation | Fujitsu Limited |
| Filing Jurisdictions | WO (2022), US (2024) | US (2022, 2024), EP (2022) |
| Core Mechanism | Live IoT data from plant assets feeds cloud-hosted twin that configures dynamically; simulation outputs propagate as control instructions back to physical asset | Clone twins temporally offset from source twin via data stream synthesizer node; concurrent clones generate forecasts of future system states |
| Primary Problem Addressed | Closed-loop control and on-demand simulation access for industrial plant operators | Temporal forecasting of future system states without halting real-time operation |
| Technology Cluster | Closed-Loop IoT Simulation with Cloud Integration | Predictive Clone Twin Sequencing |
| Update Mechanism | Dynamic self-configuration from live IoT data; ML-guided visual inspection for model updates (2026 filing) | Data stream synthesizer node applies time increment to source twin output for each clone |
| Multi-Jurisdiction Family | Yes — WO and US filings for same system | Yes — only multi-jurisdiction, multi-year family in dataset targeting temporal forecasting |
| Most Recent Filing | US, 2026 (automated ML synchronization) | US, 2024 (continuation of clone sequencing family) |
Frequently Asked Questions: Digital Twin Real-Time Process Simulation
According to multiple literature reviews in the retrieved dataset, the core architecture comprises four elements: a physical entity instrumented with sensors, a data pipeline, a virtual model executing simulation logic, and a synchronization layer managing update frequency and fidelity. This structure is described formally in the synchronization problem framework retrieved from the dataset.
Predictive clone twin sequencing involves creating one or more clone digital twins that are temporally offset from a source twin. The source twin operates in real time ingesting live IoT sensor data; each clone twin is driven by the source twin’s output with an added time increment via a data stream synthesizer node. By running clones concurrently, the system generates forecasts of future system states without halting real-time operation. Fujitsu Limited holds the key patent family for this approach across US and EP jurisdictions (filings from 2022 to 2024).
In this dataset, Siemens entities collectively account for 5 filing entries (the highest count in retrieved records), spanning WO and US jurisdictions from 2022 to 2026. Fujitsu Limited and ETRI each account for 4 filing entries. Hitachi, Ltd., PwC Product Sales LLC, IBM Corporation, SmartRac Technology Fletcher, and Satavia Limited each account for 2 filing entries in retrieved records.
The synchronization problem refers to the challenge of keeping the virtual model aligned with evolving physical reality — specifically managing both the frequency and mechanism of model updates. A formal academic treatment of this problem appears as a 2023 literature record in the dataset titled ‘The digital twin synchronization problem: Framework, formulations, and analysis.’ Siemens Corporation’s 2026 US patent on automated ML-guided synchronization addresses this via ML models that detect spatio-temporal differences between as-built and as-planned models.
Chaos Industries (US, 2025) claims a system using generative transformer networks and large language models to generate digital twins with realistic simulation of real-world conditions. Rather than manually engineering simulation models — historically a months-long effort — LLMs are used to synthesize the model structure from available data. This is identified in the dataset as a 2024–2026 directional signal under the maturation phase of DT-RTPS innovation.
The retrieved dataset includes filings covering: underground infrastructure (China University of Mining and Technology-Beijing utility tunnel, US 2025), cleanroom environments (Locus Cell Co., US 2026), 6G telecommunications network management (BTS Corporate Information Technologies, WO 2024), financial IT infrastructure resilience (Bank of America Corporation, US 2024), and live events and immersive entertainment (Meta Live Inc., WO 2024 and IN 2026). This signals that DT-RTPS infrastructure is becoming horizontal technology across sectors.
Data and insights on this page are based on a limited patent and literature dataset and are for reference only. Figures may not represent the complete technology landscape.