Atos has worked with National Grid to deliver an award-winning digital twin and data visualisation tool designed to accelerate electricity network planning and support the UK’s decarbonisation goals.

The digital twin, known as Triton, creates a digital replica of National Grid’s physical infrastructure and brings together thousands of datasets from across the network. By combining these data sources into a single environment, Triton enables National Grid to model future electricity demand, test network scenarios and make faster decisions on where grid reinforcements are required.

National Grid, which operates at the centre of the energy system in the UK and US, is responsible for moving energy from where it is generated to where it is needed. As demand rises in line with electrification and decarbonisation targets, the ability to plan network expansion more efficiently has become increasingly critical.

Owen Wilkes, network design director at National Grid, said, “Triton demonstrates how innovation, data and technology can drive positive change and accelerate progress towards our decarbonisation goals. Through our collaboration with Atos we have created a product that will enable National Grid to virtually model scenarios for how, when and where we expand our network and services to meet increased demand for energy, ensuring we continue to deliver a resilient, future-ready electricity network.”

According to the companies, Triton will “enable National Grid to accelerate how it plans the future electricity network and support the significant increase in energy demand required to achieve Great Britain’s decarbonisation goals”. By designing the platform to process and manage large volumes of network data, National Grid can reduce the time needed to analyse and decide where to reinforce the network by 70%.

Graham Scanlon, head of critical national infrastructure for Atos in the UK and Ireland, said, “This tool will improve efficiency and remove risk for National Grid. They can now make smart data-based decisions quickly. We’re pleased to have been able to support both on their current operations and needs but also on their future business. We are continually innovating on how to use technology to meet the challenge of climate change and shifting geo-political winds.”

Atos, a global IT and digital transformation firm with around 63,000 employees and annual revenue of €8 billion, has undergone significant restructuring over the past year.

FourKites has launched Loft, an AI orchestration platform designed to turn enterprise decision logic into automated workflows that run across internal systems while drawing on real-world external intelligence.

At the centre of Loft is Sophie, an AI developer agent that translates operational requirements described in natural language into production-ready automations. Rather than layering AI on fragmented systems, the platform is built to orchestrate actions across ERP, TMS, WMS, ITSM and CRM environments, while incorporating live data from the FourKites Intelligent Network.

That network spans more than 500,000 trading partners across 176 countries and processes over three million daily supply chain events. By combining this external intelligence with internal enterprise data, Loft aims to move organisations beyond dashboards and into system-level execution.

“Most enterprise operations still run on fragmented systems held together by spreadsheets, shared inboxes, and email chains,” said Josh Jewett, operating partner at NewRoad Capital Partners and former CIO of Dollar Tree and Family Dollar. “In that environment, critical decisions don’t live in systems at all — they live in Slack threads and people’s heads. When AI is layered on top of that fragmentation, it can observe problems but can’t reliably act on them.”

Loft introduces Agent Operating Procedures (AOPs), which record not only what actions agents take, but why decisions were made and who approved them. This creates a reusable knowledge layer that sits alongside the company’s digital twin of orders, shipments, inventory and assets.

“We didn’t build AI features on top of legacy software. We built an AI-native system from the ground up,” said Mathew Elenjickal, founder and CEO of FourKites. “When our AI agents do the work, we record how decisions were made — not just what happened, but the context, the prior cases that informed it, and who approved it.”

Loft is built on the company’s Intelligent Control Tower architecture, which combines network data, digital twins and a digital workforce. According to Charles Brennan, Senior Analyst at Nucleus Research, “FourKites’ Intelligent Network provides the external, real-time data foundation that AI and automation depend on.”

By embedding that external context into governed workflows, FourKites positions Loft as a platform for scaling digital twin-driven automation beyond supply chain visibility and into enterprise-wide orchestration.

Dexory has launched a next-generation autonomous warehouse robot alongside a new software capability, Storage Health, strengthening the data foundations of its digital twin platform, DexoryView.

The company’s robots have been operating in live environments since 2023, capturing real-time operational data across warehouse facilities. The new design builds on that deployment base with faster performance and a wider scanning range of up to 60 feet, compared with 40 feet in the previous generation. This extended reach allows the system to create a more complete digital replica of complex storage environments, including double-deep racking, block storage and non-racked areas.

The robot continuously captures high-frequency data and feeds it into DexoryView, which provides a live digital twin of warehouse operations. According to Richard Williams, VP of Robotics at Dexory, the aim is to ensure that operational systems mirror physical reality as closely as possible.

“Warehouse performance depends on how closely operational systems reflect reality on the floor,” he said. “By continuously capturing accurate data across every storage type and operational area, our next-generation solution gives customers a trusted, real-time foundation for decision-making. This enables warehouse operators to make decisions based on what is actually there, not what they assume is true.”

Dexory says customers using its existing systems have reported an 80% reduction in audit time and a 20% improvement in throughput. The updated robot is designed to extend those gains, while its modular architecture allows additional capabilities, such as pick face analysis or temperature monitoring, to be integrated without major hardware changes.

Alongside the hardware upgrade, Storage Health introduces an AI-driven inspection layer within the digital twin workflow. Running automatically during scans, it analyses high-resolution imagery to flag issues such as damaged racking, unstable pallets, hanging shrink wrap and crushed goods.

“The biggest risks often sit higher up or deeper in the racks where manual checks are infrequent and ineffective,” said Chris Coote, Director of Product at Dexory. “Storage Health changes this, enabling operations, health and safety, and inventory teams to act early and reduce risks before they escalate into costly incidents, injuries, or compliance issues.”

The dual launch positions the digital twin not only as a visibility tool, but as an operational risk management layer embedded directly into day-to-day warehouse workflows.

NetBrain Technologies has introduced a new version of its platform centred on what it calls Agentic NetOps, embedding agentic AI into core network operations workflows and building on its existing digital twin foundation.

The Burlington-based company says the release enables NetOps teams to autonomously investigate and diagnose complex network issues, suggest fixes and support engineers in executing remediations across hybrid-cloud environments.

“With this release, we’ve delivered agentic AI as a full-fledged digital engineer – it can diagnose complex issues independently dramatically improving the speed that teams can resolve incidents driving to the ultimate goal of preventing network downtime,” said Song Pang, Chief Technology Officer at NetBrain. “We understand our customers are at various stages of AI adoption. Our goal is to meet them where they are and make available in-the-loop, on-the-loop, and out-of-the-loop in support of our goal of reducing outages for our customers by 50% this year, and every year thereafter.”

Central to the update is the use of NetBrain’s digital twin and intent-based automation to power AI-driven diagnosis. The AI Deep Diagnosis feature operates in a human-on-the-loop mode, analysing network issues and presenting step-by-step reasoning with visualised root-cause results on a live map. The company says this approach helps junior engineers work more effectively while reducing guesswork in complex troubleshooting scenarios.

An AI Runbook Companion supports human-in-the-loop workflows, building and recommending runbooks while keeping engineers in control of approvals and execution. Additional capabilities include extended cloud support across more than 200 services, full Kubernetes integration, automated validation checks and pre-approved change remediation workflows.

Barbara Scarcella, Chief Customer Officer at NetBrain, said: “We believe customers should expect to improve the availability of their networks every year, reducing outages 50%, year over year – even as networks become more complex.”

Industry analysts see this as part of a broader shift. “We’re now moving into the phase where AI in networking must prove its value through tangible productivity gains—and that’s exactly what’s starting to happen,” said Jim Frey of Omdia. “Once organizations build trust and integrate AI into their daily workflows, they’ll see measurable improvements: faster outage resolution, proactive issue prevention, and a real opportunity to keep the toil factor under control.”

MIPS, a GlobalFoundries company, will use its platform at Chiplet Summit 2026 to underline how digital twin modelling is reshaping the design of safety-critical physical AI systems.

The company will present across multiple sessions at the Santa Clara Convention Center event on 17–19 February. Drew Barbier, VP Products at MIPS, will deliver a talk titled “Safety-Critical Physical AI Applications Using RISC-V” and join a panel discussion on selecting chiplets for edge applications. Sam Grove, Head of Software and Tools, will present “Software Guided Intelligence for Accelerating Chiplet Designs”.

While the agenda spans RISC-V and chiplet-based architectures, MIPS’ messaging places particular emphasis on digital twin-driven development. Grove said: “A software-first approach is the foundation of modern system design. At MIPS, we are enabling a ‘shift-left’ methodology through digital twin platform modeling with the MIPS Atlas Explorer, delivering data-driven insights into workload behavior and pre-silicon validation. As the demand for high-performance, domain-specific compute accelerates, the ability to analyze and optimize interactions between workloads and customizable compute platforms becomes a true competitive advantage.”

This shift-left approach, built around the MIPS Atlas portfolio, is designed to allow engineers to model processor, workload and chiplet interactions before hardware is committed. In safety-critical automotive, industrial and embedded AI contexts, pre-silicon validation and workload analysis through a digital twin can reduce integration risk and compress development cycles.

Barbier said: “We look forward to attending the 2026 Chiplet Summit as it brings key players in the semiconductor ecosystem, as well as serves as the ideal venue to discuss how RISC-V processors and software are critical to the next generation of physical AI applications. Our sessions will explore software-first platform design and how chiplets enable workload-specific customization, leveraging the MIPS Atlas portfolio.”

According to Chuck Sobey, General Chair of Chiplet Summit, “Processor decisions play a key role in system design. MIPS’ leadership will offer our attendees the insights they need to adopt software-first RISC-V designs across a wide array of applications.”

As chiplet architectures gain traction in automotive and industrial markets, the integration of digital twin platform modelling into processor selection and system validation appears set to become a defining feature of next-generation physical AI development.

The global smart factory market is projected to grow from USD 132.08 billion in 2025 to USD 271.98 billion by 2032, according to a new analysis from Maximize Market Research. The report forecasts a compound annual growth rate of 10.87% over the period, driven by the expanding integration of AI, industrial IoT (IIoT), robotics, digital twins and cloud-based analytics across manufacturing sectors.

At the core of this growth is the increasing deployment of digital twin and simulation technologies. The study highlights virtual representations of physical assets, real-time performance monitoring and predictive scenario modelling as central enablers of next-generation production. Hybrid twins that combine physics-based models with data-driven analytics are expected to play a growing role in optimising processes and anticipating supply chain disruptions.

Industrial robotics remains the leading product segment, supported by demand for precision manufacturing, continuous production and labour optimisation. AI and machine learning solutions are also gaining ground, particularly in predictive maintenance, defect detection and production planning. IIoT platforms continue to underpin connected manufacturing ecosystems, enabling machine-to-machine communication and real-time analytics.

By end-user industry, automotive manufacturers are at the forefront of adoption, integrating AI-driven assembly lines, electric vehicle production systems and digital twin technologies. Electronics, aerospace and defence, pharmaceuticals and industrial equipment sectors are also investing heavily in automation and data-driven operations.

Regionally, Asia Pacific accounts for more than 41% of the global market, with China, Japan and South Korea accelerating Industry 4.0 deployment. North America ranks second, with manufacturers retrofitting legacy plants and embedding AI-enabled predictive maintenance and connected production systems.

Recent developments cited in the report include expanded digital twin and AI analytics capabilities from Siemens AG, as well as cloud-automation collaborations between Rockwell Automation and Amazon Web Services.

The analysis concludes that the sector is entering a decisive expansion phase, with solution providers focused on interoperable platforms, cybersecurity and scalable analytics expected to benefit most as manufacturers pursue intelligent, energy-efficient production models.

Vancouver-based Industrio has formally launched Signal, a secure, enterprise-grade operational intelligence platform developed through multiple customer deployments in data-rich, mission-critical environments.

Founded in 2023 and already profitable, Industrio positions Signal as a response to fragmented data, disconnected workflows and the limits of narrow AI tools. Rather than entering the market as a conceptual product, the platform has evolved through three years of customer engagements in settings where governance, security and operational reliability were central requirements.

“Today’s announcement marks the culmination of Industrio’s customer-driven mission to solve real operational challenges with responsible AI-driven technology that works not by replacing people, but by working with them to achieve mission-critical performance,” said Edoardo De Martin, Co-Founder and CEO of Industrio. “With Signal already in use across systems where failure carries real-world consequences, our platform is grounded in proof, not promises. Instead of chasing hype, we’ve been working alongside actual customers to understand their needs in order to build secure and scalable operational systems that empower their teams.”

Among its collaborators is Pacific Blue Cross, where Industrio designed and implemented a digital twin solution processing 150,000 annual healthcare applications. The dual-layer AI platform models the full enrollment ecosystem while integrating disparate systems into a unified operational view.

At the City of Vancouver, Industrio is developing a Common Operating Platform with digital twin and AI capabilities to support major event operations. With Unilia Fuel Cells, the company has automated data ingestion from test stations to create a scalable DataOps architecture and custom reporting dashboards.

Signal is designed for environments where data is distributed across systems, decisions must remain accountable and explainable, and human expertise is augmented rather than replaced. Core capabilities include automated data ingestion, event-driven workflows, domain-specialised AI agents and a real-time operational model that evolves as conditions change.

“As a platform that is 100-percent Canadian-built, Canadian-owned and enterprise secure, Signal is an ideal candidate for dual-use applications,” De Martin said. “With Canada urgently needing technology that unlocks these applications across defence and national infrastructure, Signal is the right platform at the right time.”

Arash Ashtiani, Industrio Co-Founder and Head of AI, added: “Leaders across sectors are becoming more precise about what value means in their operations, and how data drives it. Unlike off-the-shelf tools confined to narrow tasks, Signal is purpose-built to support Canada’s critical systems. Beyond all the AI hype, Canada and the world need operational intelligence that teams can trust, that’s secure by design, and that has been proven in real practice.”

GeoConvergence will publicly demonstrate its Scan2Twin methodology for the first time at Geo Week 2026, outlining a practical approach to creating operational digital twins for large facility portfolios. The live demonstration is positioned as a response to the cost, complexity and scalability challenges that have limited wider digital twin adoption beyond pilot projects and visualisation-led use cases.

The demonstration will take place on Sunday, 16 February 2026, from 3–4pm in the Demo Zone (Booth 535) at the Colorado Convention Center in Denver.

Scan2Twin is described as an Esri-native workflow that controls the full pipeline from reality capture through BIM processing to deployment within ArcGIS Indoors. Rather than focusing on visual representation alone, the methodology structures data around assets, work orders and space usage, enabling digital twins to be used directly for day-to-day facility operations.

The approach combines multiple mobile mapping technologies to capture survey-grade data without disrupting live environments. These include wearable mobile mapping systems for continuous capture, handheld scanners for detailed environments, and a robotic quadruped platform to access restricted or low-clearance spaces. GeoConvergence also applies automated anonymisation to support PII and HIPAA compliance where required.

During the one-hour session, attendees will see raw scan data processed into an operational digital twin in near real time. The demonstration will show how facilities teams can use the resulting models for turn-by-turn indoor routing, spatially contextualised work orders, and space management and planning. The workflow covers reality capture execution, BIM processing, asset extraction and ArcGIS Indoors deployment, including GeoBIM outputs.

Alongside Scan2Twin, geoConvergence will also present ReservAssist and Scenario Planner, two proprietary tools designed to extend digital twin functionality into space reservations and facility planning activities.

The session is aimed at facility managers, GIS professionals deploying ArcGIS Indoors, and organisations responsible for managing large or distributed facility portfolios. geoConvergence’s technical team will be available throughout the conference to discuss workflows, applications and implementation considerations.

Founded in 2004, geoConvergence is an Esri Gold Business Partner specialising in digital twin solutions for facility management. The company focuses on Esri-native workflows that integrate mobile LiDAR capture, BIM processing and indoor GIS deployment without requiring custom development, aligning with existing enterprise GIS environments.

Edge Total Intelligence Inc., a provider of real-time digital operations solutions, is set to demonstrate its edgeCore AI-powered digital twin and secure Data Mesh platform at the 2026 WEST Conference in San Diego, taking place February 10–12. The event, focused on maritime and defence technology, draws senior leaders, engineers, and operators from the U.S. Navy, Marine Corps, Coast Guard and associated industry partners.

EdgeTI plans to showcase the edgeCore platform — described in the press release as having achieved Technology Readiness Level 9, the highest maturity level — at Booth 1152. According to the company, the platform brings together digital twin capabilities with real-time data integration and orchestrated AI, aiming to help organisations unify siloed systems, accelerate AI and machine-learning adoption and support faster decision-making without costly data replication into traditional repositories.

EdgeTI’s Chief Revenue Officer, Jacques Jarman, said the company is “excited to participate in WEST 2026 to showcase our secure AI-powered Digital Twin capabilities with the AFCEA community.” He noted the shifting strategic landscape and the need for tools that let operators “turn data into actionable decision-ready knowledge at the speed of relevance.” Jarman added that edgeTI aims to provide commanders with “the capability to achieve data dominance at a fraction of the cost of legacy tools within weeks.” (Stock Titan)

EdgeTI positions its edgeCore technology as a commercial‐off-the-shelf (COTS), patented solution designed to connect disparate systems and data sources in real time, eliminating redundant storage and enabling a purpose-built data mesh. The company asserts that this approach helps overcome long-standing challenges in complex, mission-critical environments by delivering actionable insights where data already lives rather than moving it into separate warehouses or lakes.

The WEST Conference, organised by AFCEA International and the U.S. Naval Institute, serves as a major forum for discussing sustainment, shipbuilding, readiness and future warfighting capabilities. edgeTI’s presence reflects wider industry interest in digital twin technologies, particularly for enhancing situational awareness, operational agility and lifecycle management across defence and industrial domains.

Iveda, a provider of AI video analytics and smart city solutions, has introduced a new Traffic Management Module within its IvedaAI™ platform to expand its role in urban mobility and deliver fresh revenue opportunities for municipalities and technology partners. The launch is aimed at transforming how cities use existing intersection camera networks to manage traffic more intelligently and cost-effectively.

The module turns video feeds from current infrastructure into real-time traffic intelligence, offering capabilities such as vehicle counting, speed monitoring, congestion insights and automated detection of safety events like red-light violations and collisions. By doing so, it seeks to help cities improve traffic flow, enhance roadway safety and support faster emergency response times - all without requiring investment in new roadside sensors or hardware.

According to Iveda, the traffic management segment represents a significant growth opportunity within the company’s smart city roadmap. The company reports active engagements with multiple U.S. cities and transportation agencies for pilot deployments in 2026 and strong interest from its reseller network to bring the new capability to municipal and transport customers.

“Imagine all the cameras you already see above intersections across America,” said David Ly, CEO of Iveda. “Today, most are limited to basic signal optimization. With IvedaAI, those same cameras can instantly become tools for smarter traffic flow, faster alerts, and safer streets — without requiring cities to replace infrastructure. That simplicity is where the value is.”

Industry analysts project that the intelligent traffic management systems market could grow to nearly $28 billion by 2030 as cities accelerate adoption of AI-driven transportation technologies. The new module’s design emphasises ease of deployment, with capabilities such as a digital twin-style interface and scalable software that can be set up quickly. According to Ly, the implementation cost is roughly the equivalent of $20,000 per intersection, representing a potentially high return on investment for cities looking to modernise traffic analytics without major infrastructure upgrades.

This launch is also positioned as a strategic step in productising IvedaAI’s broad video intelligence capabilities, marking one of the company’s first packaged solutions focused on a single high-demand vertical and generating recurring software revenue alongside its existing platform offerings.

Valeo and Anritsu have announced a collaboration aimed at accelerating the virtual validation of software-defined vehicles, using digital twin technology to support faster and more scalable development of connected automotive systems.

The partnership focuses on enabling virtual validation of vehicle software domains such as telematics units, an area growing in importance as automakers shift toward software-defined architectures. The two companies will present a live demonstration of their joint approach at MWC 2026 in Barcelona, running from 2–5 March 2026.

As vehicles become increasingly connected to cloud infrastructures, cellular connectivity is emerging as a critical component of software-defined vehicle platforms. This shift brings new testing challenges, including the need to validate performance across diverse networks, operating conditions, and real-world edge cases. Traditional validation approaches rely heavily on hardware-based testing and manual parameter extraction, often requiring complex test benches that struggle to keep pace with modern continuous integration and continuous deployment (CI/CD) cycles.

To address this, Anritsu is integrating a virtual engine into its software-based In-the-Loop Simulator, capable of reproducing real-world field events. This enables the creation of a digital twin of complex, multi-network environments, including inter-mobile network operator scenarios and vehicle-to-network and vehicle-to-vehicle communications. These simulations can be run in cloud environments or on local machines, supporting the virtual testing of multiple vehicles in parallel.

Gilles Mabire, CTO of Valeo’s Brain division, said: “Telematic unit and connectivity in general is a strategic pillar of software-defined architecture. This product, developed by Valeo, has to connect and operate in all cellular networks and in various and diverse field conditions. The growing complexity of development happens in the context of reduction of vehicle development cycle and is associated with a continuous reduction of vehicle cost and efforts. Combining virtual telematics unit and virtual cellular network running both in the cloud is an innovative solution to tackle these challenges and is the baseline of this partnership with Anritsu.”

Yukiharu Ogawa, General Manager of Anritsu’s IoT Test Solution Division, added: “To ensure stable utilization of sophisticated connectivity services, it is essential to implement software that takes into account the edge cases that occur in the real world. Anritsu addresses this challenge with a one-stop toolchain that comprehends everything from field driving tests to simulations in the lab.”

According to Ogawa, integrating Anritsu’s testing capabilities with Valeo’s software development environment allows car manufacturers to carry out end-to-end verification, from in-vehicle systems through to backend services, within a fully virtualised digital twin environment.

A group of researchers has demonstrated how combining building information modelling (BIM) with life cycle assessment (LCA) can create a comprehensive digital representation of a building’s carbon emissions from design through demolition, an advance with broader implications for digital twin applications in complex engineered systems.

In a study published in Energy & Environment Nexus, Xinyu Yang and colleagues used an office building in Taiyuan, China, as a case study to link a 3D BIM model with a carbon emission estimation tool (the Carbon Emission Estimator for Buildings, CEEB). The integrated model quantifies greenhouse gas emissions across four principal stages of a building’s life cycle: production and transport of materials, construction, operation and maintenance, and demolition.

The results show that operational energy use - heating, cooling, ventilation, and equipment - accounts for the overwhelming share of total emissions, roughly 94–95 per cent of the life-cycle footprint. Emissions from material production and transport were responsible for around 9–10 per cent, with steel, concrete and cement dominating that category. Construction and demolition together contributed only around 0.1 per cent, making them comparatively minor in the overall carbon picture.

By integrating LCA into the BIM environment at the design stage, the researchers were able to highlight where carbon intensity is concentrated before physical construction begins. This early visibility allows designers to test alternatives and optimise decisions where they matter most, rather than retrofitting carbon-reduction measures after the fact.

“By embedding life cycle carbon data within the BIM process, we can forecast the environmental consequences of design choices and prioritise efficiency in operation,” the authors write.

Although the study focuses on the built environment, the approach resonates with current efforts in automotive development to build digital twins that capture real-world performance metrics over a system’s life span. In complex products like vehicles, early integration of emissions and energy models with design data can similarly help engineers identify where design choices drive the largest environmental impacts.

The research underscores the value of digital twins not just as tools for visualisation but as platforms for embedding multi-domain simulation and assessment. As industries pursue net-zero goals, linking detailed digital models with environmental analytics could become an essential capability for both infrastructure and automotive sectors.

Nvidia and Dassault Systèmes have deepened their collaboration around digital twins and artificial intelligence, positioning what they describe as “industry world models” as a foundation for simulating, testing and optimising complex industrial systems before changes are made in the physical world.

The partnership brings together Dassault Systèmes’ long-established 3DExperience platform with Nvidia’s AI and accelerated computing stack, including Omniverse, to create high-fidelity virtual representations of factories, vehicles, infrastructure and products that can be continuously updated with real-world data. The goal is to move beyond static digital twins towards living models that can reason, predict and respond as conditions change.

According to the companies, these industry world models are designed to capture not only geometry, but also physics, behaviour and operational constraints. By combining physics-based simulation with generative and predictive AI, engineers and operators can explore scenarios such as production line changes, supply chain disruptions or new product designs in a virtual environment, reducing risk and shortening development cycles.

In manufacturing, this approach could allow teams to simulate entire plants, testing layout changes or automation strategies before committing capital expenditure. In automotive and aerospace development, it enables more detailed virtual validation of systems performance, energy efficiency and safety, while also supporting collaboration across engineering disciplines. The same principles extend to energy, infrastructure and other asset-intensive industries where decisions are costly and difficult to reverse.

Nvidia’s role is not simply to provide raw compute, but to enable AI models that can be trained within these digital environments. By learning from vast amounts of simulated and real data, AI agents can help identify inefficiencies, predict failures or suggest optimised operating parameters. Dassault Systèmes, meanwhile, contributes decades of domain-specific modelling expertise and workflows already embedded in industrial organisations.

This convergence reflects a broader shift in how digital twins are being used. Rather than standalone engineering tools, they are increasingly becoming strategic platforms that link design, simulation and operations. For organisations adopting digital twins at scale, the combination of trusted physics models with AI-driven insight points towards more autonomous, adaptive systems.

As industries face mounting pressure to improve efficiency, resilience and sustainability, the ability to test decisions safely in a virtual world before acting in the real one is becoming a competitive necessity rather than an experimental technology.

Source: cio.com / Lynn Greiner

Applied Computing has been named a recipient of Frost & Sullivan’s Global Transformational Innovation Leadership Award for Energy AI, with the recognition closely tied to the company’s role in advancing AI models that operate alongside industrial digital twins.

The London-based company has developed Orbital, a foundation AI model designed for complex energy environments such as refineries, petrochemical facilities and renewable energy assets. Rather than acting as a standalone analytics tool, Orbital is positioned to work within live operational systems and digital twin environments, using physics-grounded intelligence to mirror, predict and optimise real-world behaviour.

By combining predictive modelling, natural language interfaces and real-time optimisation, Orbital supports a core promise of digital twins: enabling operators to test decisions, understand system behaviour and intervene proactively without disrupting physical operations. In practice, this allows energy companies to move from static or descriptive digital twins toward continuously learning, decision-ready twins that can adapt to changing conditions.

“This recognition reflects the scale of the challenge energy operators face and the importance of solutions that work in real-world conditions,” said Callum Adamson, CEO of Applied Computing. “Orbital was built to operate inside the most complex industrial environments on the planet, and this award validates the progress we are making in bringing superintelligent, physics-grounded AI into production.”

Frost & Sullivan’s assessment focused on Applied Computing’s ability to shift energy operations from reactive, manual processes to autonomous optimisation. From a digital twin perspective, this signals a broader industry transition: twins evolving from visualisation and simulation tools into active operational systems, where AI continuously calibrates the virtual model against live data and recommends or executes actions.

The company’s recent partnerships underline this direction. Collaborations with Databricks and Kongsberg Digital are aimed at embedding Orbital into existing data platforms and industrial digital twin frameworks, strengthening the link between AI reasoning, simulation models and operational decision-making. This alignment enables digital twins to operate at scale across fleets of assets rather than as isolated engineering models.

“Applied Computing transforms the energy sector from manual, reactive operations to autonomous, AI-driven systems that optimize performance while accelerating decarbonization efforts,” said Karthik Sundaram, Research Director, Industrial, Frost & Sullivan. “Applied Computing is augmenting the energy operations solutions, creating the first foundational model built to overcome sector-specific pitfalls.”

The award comes during a period of rapid growth for Applied Computing, including a £9 million seed funding round led by Stride.VC and the appointment of former Shell AI leader Dan Jeavons as president. As energy operators face increasing pressure to improve efficiency, resilience and decarbonisation, the convergence of AI and digital twins highlighted by this recognition is emerging as a critical capability rather than an experimental add-on.

Principle, a San Francisco-based strategic foresight platform, has raised $2 million in pre-seed funding to commercialise what it describes as a new approach to corporate strategy: continuous, simulation-based decision-making using digital twins of markets, competitors and organisations.

The round was led by SMRK VC and SMOK VC, with participation from Ride Home AI Fund, a16z Scout Fund, Bain Capital Scout Fund and Unpopular Ventures.

Principle positions its platform as an answer to what it sees as a persistent enterprise blind spot: strategy processes that rely on static documents and lagging indicators. While companies such as Shell have long used scenario planning, the approach has typically required specialist teams and long planning cycles.

“Every Fortune 500 company we talk to has the same pattern,” said Artur Kiulian, co-founder and CEO of Principle. “They misread market signals, they react too slowly to structural shifts, and by the time something obvious in hindsight becomes clear, it’s already too late.”

At the core of the platform is the creation of digital twins representing a company, its competitors, regulators and broader market forces. These models are then used to run adversarial simulations across hundreds of possible strategic directions. Unlike one-off analyses, the simulations are persistent and update as new information arrives, including competitor moves, regulatory changes and M&A activity.

The system is already in pilot deployments with Fortune 500 companies, governments and energy sector organisations, as well as technology firms evaluating acquisition strategies. Oleksandr Kosovan, CEO and founder of MacPaw, said: “Principle helps us build that muscle: turning strategy into an iterative process where we explore scenarios, learn quickly, and decide based on evidence, not intuition alone.”

The company’s origins lie in public-sector work on national AI capabilities, including a memorandum of understanding with Ukraine’s Ministry of Digital Transformation and projects in the Middle East. Collaborations with the Dubai Future Foundation and the Dubai Centre for Artificial Intelligence focused on scenario modelling for health system resilience. “The future belongs to those who ask the most important questions,” said Khalfan Juma Belhoul, CEO of Dubai Future Foundation.

Principle’s founders argue that recent advances in large language models make it possible to simulate how markets and organisations behave, rather than just analysing historical data. The team brings experience from Google, CERN and strategic technology work for the White House. The platform is being developed on Amazon Web Services’ Nova architecture, following a path also taken by companies such as Reddit, Sony and Booking.com.

“Enterprise AI has optimised individual productivity,” said Yurii Filipchuk, co-founder of Principle. “What’s missing is the layer that shapes where the business actually goes. We adopted military wargaming techniques and applied them to corporate strategy.”

The funding will be used to productise the platform, expand real-time market intelligence integration and grow enterprise pilots across industrial, technology and financial services sectors.

Siemens has acquired Grenoble-based Canopus AI, extending its electronic design automation (EDA) portfolio with AI-driven computational metrology and inspection capabilities aimed at advanced semiconductor manufacturing.

The acquisition brings Canopus AI’s wafer and mask metrology software into Siemens’ semiconductor design and manufacturing digital thread, reinforcing the company’s push toward higher-fidelity manufacturing digital twins. As semiconductor geometries continue to shrink and production volumes increase, manufacturers face growing challenges in maintaining yield, process control, and inspection accuracy at advanced nodes.

Canopus AI’s technology focuses on what it calls “Metrospection,” an AI-enabled approach that combines conventional wafer metrology with inspection data to improve measurement accuracy and decision-making. By embedding these capabilities within Siemens’ existing EDA and manufacturing simulation tools, the combined portfolio is positioned to support more precise modeling of real-world fabrication processes.

“The acquisition of Canopus AI exemplifies Siemens’ commitment to leveraging industrial AI to solve critical challenges in semiconductor manufacturing,” said Tony Hemmelgarn, president and CEO of Siemens Digital Industry Software. “By combining the computational lithography and manufacturing physics simulation capability in our Calibre portfolio with Canopus AI’s advanced metrology and inspection technologies, we are creating a differentiated, end-to-end EDA digital thread that improves the fidelity of printed wafer patterns, accelerates yield ramp and reduces time-to-volume for advanced nodes.”

Hemmelgarn added that the integration advances Siemens’ vision of a comprehensive semiconductor manufacturing digital twin, enabling tighter process control and improved mask development at sub-nanometer scales.

Founded in 2021, Canopus AI develops software designed to bridge gaps between design intent, inspection data, and manufacturing outcomes. Its AI-driven models aim to enhance both wafer and mask inspection workflows by extracting more actionable insight from large-scale measurement data.

“We are delighted to join Siemens and bring the power of AI-enabled metrology in the semiconductor industry to a broader audience as part of Siemens’ EDA community of users,” said Joël Alanis, chief executive officer of Canopus AI. “Together, we’ll empower innovators pushing the boundaries of semiconductor design and manufacturing with robust wafer and mask metrology and inspection.”

envelio has expanded its Intelligent Grid Platform (IGP) with the launch of Strategic Grid Planner, a new application designed to support automated, load-flow–based planning of electricity distribution networks across multiple future time horizons.

The Strategic Grid Planner is built to help utilities create and validate technically sound and economically viable grid development plans for target years such as 2030, 2035, and 2045. It extends the role of the IGP’s digital twin beyond operational analysis, positioning it as a system-wide planning environment capable of evaluating thousands of potential grid configurations across defined planning periods.

As distributed energy resources, battery storage, electric vehicle charging infrastructure, and increasingly volatile load profiles place growing pressure on distribution networks, utilities are facing complex, long-term investment decisions without tools that can systematically explore the full range of feasible system options. envelio’s new planning application addresses this gap by using algorithmic optimisation, high-performance simulation, and advanced network modelling to assess alternatives at scale.

Luigi Montana, CEO of envelio Inc., said: “The current pace of change in the distribution grid is unprecedented. This results in more complexity, more uncertainty, and more volatility. System planning is a major lever as the decisions made today will affect the next 5–20 years and beyond. But today’s approach to system planning was not designed for the current complexity and volatility. A new approach is needed to plan continuously, assess thousands of alternatives, and leverage technology to find optimal solutions.”

Within the IGP environment, Strategic Grid Planner automatically generates and compares grid expansion scenarios, assessing cost and technical performance across different future states. The software evaluates solutions in a structured sequence, prioritising grid optimisation measures first, followed by flexibility options, reinforcement, and finally physical expansion. This allows planners to exhaust lower-cost operational improvements before committing to infrastructure upgrades.

At the core of the application is a self-learning, AI-based algorithm that improves with each planning cycle. By analysing previous scenarios, identifying recurring patterns, and recognising dependencies between grid measures, the system progressively refines its recommendations and predicts the quality of potential expansion strategies. Planners retain the ability to manually review and adjust proposed designs.

In the U.S. market, envelio says the tool can also support cluster studies by enabling scalable, scenario-based evaluation of multiple interconnection requests within a shared network context. This positions the IGP’s digital twin as a foundation not only for operational insight, but for defensible, long-term investment planning as grid complexity continues to accelerate.

Schneider Electric and ETAP have unveiled a new physics-based digital twin designed to bring engineering-grade power system simulation directly into real-time grid operations. The announcement was made at DTECH, a major US transmission and distribution industry event.

The joint solution is intended to help utilities and operators of critical infrastructure improve resilience and reduce time-to-power by unifying grid planning and operational decision-making. Rather than relying on separate models for engineering studies and live operations, the digital twin maintains a single, continuously updated representation of the electrical network that can be used across the asset lifecycle.

According to ETAP, the platform is grounded in electrical physics rather than abstract visualisation. This allows utilities to simulate switching operations before execution, validate protection schemes, and anticipate fault conditions under live grid conditions. The model can also be used for arc-flash analysis and compliance studies, supporting both operational safety and regulatory requirements.

“Until now, utilities have operated two separate worlds, one for planning, another for operations. We’ve collapsed that divide,” said Tanuj Khandelwal, CEO of ETAP. “This isn’t simulation anymore. It’s a living digital twin that thinks alongside the grid while validating protection schemes before they execute, anticipating faults before they cascade.”

The digital twin is integrated with Schneider Electric’s One Digital Grid Platform and EcoStruxure ArcFM Web, linking geospatial asset data with real-time operational data. By combining GIS-based network visibility with simulation-grade models, the companies say utilities can move from static representations of the grid to predictive operational insights.

ETAP, which was acquired by Schneider Electric in 2021, provides power system modelling and analysis software used across utilities and mission-critical environments such as data centres, healthcare facilities, and aerospace infrastructure. The combined offering is positioned to support grids facing increasing complexity from electrification, distributed energy resources, and extreme weather events.

“By combining ETAP’s engineering rigor with Schneider Electric’s industry-leading geospatial technologies, operators gain a unified, lifecycle model that reduces risk and accelerates modernization,” said Ruben Llanes, CEO, Digital Grid, Schneider Electric.

Schneider Electric reports that ETAP technology has been validated across more than 50,000 installations globally, including high-availability and safety-critical environments, and is now being extended into live grid operations through this digital twin approach.

The European Commission has confirmed the third phase of the Destination Earth (DestinE) initiative, marking a significant step forward in the development and operationalisation of large-scale digital twins of the Earth’s systems. The programme, implemented by ECMWF (the European Centre for Medium-Range Weather Forcasting) together with ESA and EUMETSAT, aims to deliver high-fidelity, data-driven digital twins that can support decision-making around climate adaptation, risk management and long-term sustainability.

Phase three moves the initiative beyond initial development and early demonstration, with a focus on scaling, integration and user uptake. According to ECMWF, the next stage will concentrate on enhancing the scientific capabilities of the existing digital twins, improving their accuracy and resolution, and expanding access for policymakers, researchers and industry users across Europe.

Two core digital twins sit at the heart of DestinE: one focused on climate change adaptation, and the other on weather-induced extremes. These models combine observations, advanced numerical modelling and high-performance computing to simulate complex Earth system interactions, enabling scenario testing and forward-looking analysis that would not be possible with traditional forecasting tools alone.

In the announcement, Florence Rabier, Director-General of ECMWF, said: “The confirmation of the third phase of Destination Earth is a strong endorsement of Europe’s ambition to lead in digital twin technology for climate and environmental intelligence.” She added: “It allows us to move from development into a phase where these digital twins can increasingly support real-world decisions.”

For ECMWF, the third phase also reinforces its role as the computational and modelling backbone of the initiative. The organisation will continue to evolve its digital twin engine, integrating new data sources, improving model coupling and ensuring that the system can run efficiently on Europe’s emerging exascale computing infrastructure.

While DestinE is primarily framed around climate and environmental resilience, its progress has wider implications for digital twin development across sectors. The programme demonstrates how complex, multi-physics digital twins can be built, validated and operated at continental scale, offering lessons for industries exploring digital twins for infrastructure planning, energy systems and advanced manufacturing.

As DestinE enters its third phase, the emphasis shifts from proving technical feasibility to delivering actionable intelligence. For Europe’s digital twin ecosystem, this transition may prove just as important as the underlying technology itself.

GE Appliances has launched its first-ever in-house water filter manufacturing operation at its global headquarters in Appliance Park, Louisville, marking a significant expansion of its domestic production footprint as part of a broader multi-billion-dollar investment in U.S. manufacturing. The initiative reflects the company’s drive to strengthen quality, supply-chain resilience and consumer protection, while also addressing the growing problem of counterfeit filters that can compromise water quality and appliance performance.

Central to the development of this new advanced manufacturing line was the use of virtual reality and digital twin technology. By creating detailed virtual models of the equipment and facility layout, GE Appliances’ engineers were able to simulate and test performance, workflows and interactions long before physical installation took place. This approach allowed them to optimise the design, identify potential issues and refine control systems in a fully virtual environment, reducing the need for costly real-world iterations and downtime.

Digital twin tools were paired with other advanced simulation techniques and virtual-reality environments during the design phase. Together, these digital twins served as a bridge between planning and execution, enabling teams to visualise complex processes, anticipate bottlenecks and ensure the new line would integrate seamlessly into Appliance Park’s broader manufacturing ecosystem. This represents a clear example of how digital twin applications are moving beyond prototyping into full-scale industrial deployment in consumer goods production.

The physical line itself incorporates high levels of automation, including fully automated assembly and packaging, machine-vision-guided robotic pick-and-place systems, and 100 % in-line air leak testing with end-to-end traceability. Operators and technicians spent weeks onsite with the equipment manufacturer before launch, reinforcing how digital-first planning complements workforce training and operational readiness.

By combining digital twin-driven design with cutting-edge automation, GE Appliances aims to accelerate production, improve quality and enhance traceability for water filters produced in Louisville. The company says this integrated approach will help it meet consumer demand more rapidly while maintaining higher levels of consistency and performance than have been typical in outsourced filter production.

“Water quality is a concern for close to two-thirds of U.S. consumers, which is why GE Appliances has invested millions of dollars in our ice, water and filtration research and testing labs here in Kentucky,” said James Downey, senior director of Water Filtration for GE Appliances. “That commitment to quality starts with a talented research and development team and world-class laboratories, and now our new multi-million-dollar water filter line and highly skilled manufacturing team. We now own the entire process, which will give consumers greater confidence they’re getting the genuine, certified filters they expect.”

The broader initiative reflects GE Appliances’ strategy of leveraging advanced digital tools across its U.S. manufacturing network to create agile, efficient production systems that can adapt to changing market needs and drive continuous improvement.

“This investment is exactly what advanced manufacturing should look like in America,” added Bill Good, vice president of Manufacturing for GE Appliances. “We’re combining cutting-edge automation, digital tools and highly skilled teams to build a manufacturing operation that raises the bar on quality and efficiency. This new water filter manufacturing is another proof point of how we are putting our $3 billion U.S. investment commitment to work.”