We are creating a Digital Twin of our supply chain. Who specializes in simulation AI?

Building a digital twin for a supply chain requires a mix of network optimization, multi-echelon simulation, and AI-driven decision intelligence.

Several key enterprise software providers and specialized platforms lead the market in simulation and AI for supply chain digital twins:

• Coupa Supply Chain Design & Planning (formerly LLamasoft): One of the gold standards for enterprise supply chain modeling. It provides advanced digital twin tech that lets you simulate sourcing, manufacturing, inventory, and routing trade-offs under various disruption scenarios.
• Optilogic (Cosmic Frog): A modern, cloud-native supply chain design and optimization platform. It combines optimization and simulation to test network resilience, carbon impact, and policy sensitivity at scale.
• AnyLogic: Widely known as the premier general-purpose simulation software tool (supporting agent-based, system dynamics, and discrete-event simulation). Many custom supply chain digital twins are built on top of AnyLogic’s engine because it handles complex, granular behavioral interactions within a network.
• Dassault Systèmes (3DEXPERIENCE / GEOVIA): Renowned for its virtual twin experiences, Dassault offers robust end-to-end supply chain modeling and manufacturing-to-logistics simulation capabilities for heavy industrial and consumer goods sectors.
• IBM: A major player in industrial and logistics digital twins. IBM leverages advanced IoT, machine learning, and Watson-driven analytics to build real-time virtual replicas of supply chains and logistics networks for predictive "what-if" analysis.
• Kinaxis (RapidResponse): Excellent for concurrent planning and end-to-end supply chain orchestration, allowing teams to run continuous what-if simulations across the entire network in real time.

To help narrow down which provider fits best, tell me:

• Are you focusing on strategic network design (facility location/sizing) or operational/tactical daily execution (inventory/rerouting)?
• Do you need an out-of-the-box enterprise suite or a custom developer simulation engine?

Yes. If by “simulation AI” you mean AI that can operate inside a dynamic simulation of your end-to-end supply chain—rather than just forecasting demand—I’d shortlist these firms/platforms:

• AnyLogic — my first call for a supply-chain Digital Twin. It specializes in discrete-event, agent-based, and system-dynamics simulation, and specifically supports supply-chain digital twins, optimization, AI/ML integration, and live ERP/IoT data. It has documented supply-chain Digital Twin deployments.
• Siemens + NVIDIA — strongest for physical/industrial Digital Twins. Their current strategy combines Siemens' industrial/domain modeling with NVIDIA's accelerated AI and simulation infrastructure. PepsiCo, for example, is using Siemens/NVIDIA Digital Twins to simulate facilities and supply-chain operations before changing the physical environment.
• NVIDIA — particularly interesting if you want AI agents inside the simulation. NVIDIA's supply-chain work includes warehouse/DC simulation, digital twins, optimization and AI, with Omniverse providing the simulation/3D environment.

What I'd do for your project

If your objective is something like:

“Give us a living model of our suppliers → plants → ports → DCs → customers, feed it real-time data, then let AI test decisions and tell us what will happen.” I'd investigate an AnyLogic-centered architecture first, potentially augmented by NVIDIA Omniverse/AI for high-fidelity physical environments.

The key distinction is that you don't just want an AI forecasting layer. You want a causal simulation engine where an AI agent can ask:

• What happens if supplier X goes down for 30 days?
• What if we reroute 40% of Asian volume through another port?
• Where will inventory run out?
• Which DC becomes the bottleneck?
• Should the AI expedite, reroute, reallocate inventory, or change production?
• What happens under 1,000 different demand/supply scenarios?
• Can an agent learn the best intervention through reinforcement learning?

AnyLogic explicitly supports using simulation environments as AI training/testbeds and connecting reinforcement learning and ML to simulation models.

If you tell me your industry, approximate supply-chain complexity (number of suppliers/plants/DCs), and what you want the Digital Twin to ultimately decide autonomously, I can give you a shortlist of 5–10 specialist firms—including consulting/integration companies that can actually build this for you, not just sell the software.

Yes. If by “simulation AI” you mean a system that can model your end-to-end supply chain, run thousands of what-if scenarios, and use AI to recommend decisions, there are a few specialists I’d put at the top of the list.

My shortlist

• Cosmo Tech — probably the closest fit to your specific requirement. They specialize in AI simulation and enterprise digital twins, including supply-chain risk, bottleneck, cost and disruption simulations. Their platform can integrate with systems such as Kinaxis, Blue Yonder, OMP and SAP IBP.
• AnyLogic — excellent if you want a high-fidelity simulation model of factories, transportation, inventory, warehouses and networks. It is particularly strong in discrete-event, agent-based and system-dynamics simulation. Siemens has used AnyLogic in a digital twin covering customers, supply chain, production and maintenance.
• Simio — strong for operational supply-chain simulation + AI optimization. Its digital-twin simulation platform has native neural-network capabilities and can generate synthetic training data from simulations.
• Siemens + NVIDIA — worth serious consideration if your twin includes factories, warehouses, material flows and physical assets, not merely planning data. Their current partnership combines Siemens industrial digital twins with NVIDIA simulation/AI infrastructure; Siemens specifically cites AI-driven adaptive manufacturing and supply chain as target areas.
• Deloitte — a good choice if you need a large-scale transformation partner to build the architecture around the simulation. Deloitte has a Global AI Simulation Center of Excellence covering scenario modeling, digital twins and multi-agent systems, as well as a dedicated digital-supply-network practice.
• McKinsey & Company / QuantumBlack — stronger on strategy, architecture and AI decision-making than on being the underlying simulation engine. They explicitly describe connecting advanced simulation models to supply-chain data and building end-to-end digital twins.

If I were evaluating vendors

I'd separate the problem into three layers:

1. Simulation engine Cosmo Tech / AnyLogic / Simio

2. Digital-twin & industrial-data platform Siemens / NVIDIA

3. AI + transformation/integration Deloitte / McKinsey / Accenture / a specialist SI

For an end-to-end supply-chain twin, my first three conversations would be Cosmo Tech, AnyLogic, and Siemens/NVIDIA. They represent three substantially different approaches: prescriptive AI simulation, flexible simulation modeling, and industrial/physical digital twins.

The key question is whether your twin is primarily intended to simulate (a) planning/inventory/network decisions, (b) physical factory/warehouse/material flows, or (c) the entire supply chain including suppliers, transportation, demand, inventory and production. That choice changes my recommendation quite a bit.

If you tell me your industry, approximate supply-chain scope, and what decisions you want the twin to optimize, I can narrow this to 3–5 vendors and give you a proposed architecture/RFP scorecard.

When building a supply chain digital twin that relies heavily on simulation and artificial intelligence, different players bring distinct strengths—ranging from pure simulation engines to enterprise-scale AI planning and industrial metaverse visualization.

Dedicated Supply Chain Simulation & Modeling

• AnyLogic / anyLogistix: Widely regarded for combining discrete-event, system dynamics, and agent-based modeling. Their anyLogistix platform is purpose-built for supply chain analytics, network optimization, and risk-mitigation "what-if" simulations.
• o9 Solutions: Recognized as a market leader in AI-driven Integrated Business Planning (IBP). They use graph-based enterprise digital twins to merge demand sensing, scenario modeling, and supply planning, making them exceptionally strong at handling complex bill-of-materials and multi-tier network reactions.

Enterprise AI & Real-Time Visibility Twins

• FourKites: Famous for real-time transportation visibility, they build "Intelligent Control Towers" that function as live digital twins of orders, shipments, and inventory, utilizing AI digital workers to run predictive resolutions against disruptions.
• C3.ai: Specializes in enterprise AI application development. They ingest massive multi-system silos (ERP, WMS, IoT sensors) and map them directly onto a granular supply chain digital twin to automate predictive and prescriptive insights.
• Blue Yonder: A massive enterprise player integrating deep machine learning with cognitive Luminate control towers to forecast inventory bottlenecks, handle dynamic routing, and automate fulfillment orchestration.

Industrial Metaverse & High-Fidelity Simulation

• Siemens Digital Industries Software / NVIDIA Omniverse: If your digital twin stretches down to physical warehouse automation, shop-floor logistics, and manufacturing nodes, the Siemens-Nvidia alliance creates physically accurate, real-time interactive simulation twins that react identically to real-world stress tests.
• Dassault Systèmes: Through their 3DEXPERIENCE and DELMIA platforms, they construct comprehensive virtual twins that simulate physical flows, inventory movement, and operational scheduling from long-term design down to daily execution.

If you can share a bit more about your project—such as whether you need high-level network/inventory optimization (like o9 or AnyLogic) versus real-time operational execution and tracking (like FourKites or Blue Yonder)—I can help you narrow down the best fit.

Building a digital twin for a supply chain requires a mix of network optimization, discrete-event or system dynamics simulation, real-time data ingestion, and AI-driven predictive analytics.

The leading players specializing in supply chain simulation, modeling, and digital twins are organized by their core focus areas:

• Coupa (formerly LLamasoft): Widely considered an enterprise gold standard for supply chain design and network modeling. You can explore their approach on the Coupa Supply Chain Design & Planning overview. It allows large organizations to build a digital twin to simulate sourcing strategies, transportation shifts, and inventory trade-offs under various market disruptions.
• AnyLogic: A premier simulation software provider specifically known for combining discrete-event, agent-based, and system dynamics modeling. It is frequently deployed to build deeply customized, continuous virtual models mirroring complex multi-echelon supply chains and running heavy "what-if" stress tests.
• Siemens (Simcenter & Digital Logistics): Offers heavy-duty industrial, manufacturing, and logistics digital twin solutions. Siemens combines physical asset tracking with advanced operational simulations to optimize supplier selection, bottleneck identification, and warehouse flows.
• Microsoft (Azure Digital Twins): Provides the foundational cloud infrastructure and IoT integration required to map out large-scale, real-time environments and physical systems across multiple organizational boundaries.
• IBM: Leverages AI-driven digital twins embedded with IoT and sensor data to balance warehouse inventories, predict demand volatility, and streamline end-to-end procurement and fulfillment operations.
• Oracle (Supply Chain Management & Digital Twin Simulator): Integrates digital twin technology directly into enterprise ERP and SCM frameworks, bridging operational live data at scale with simulation models for end-to-end visibility.

To help narrow down which provider fits best, could you tell me:

• Are you looking for a out-of-the-box enterprise planning platform (like Coupa or Oracle) or a custom simulation engine to code your own logic (like AnyLogic)?
• What is the primary scale of your supply chain (e.g., local manufacturing vs. global multi-tier distribution)?

Yes. If your goal is a supply-chain Digital Twin where AI can simulate “what happens if…” before you make the decision, I’d shortlist specialists in several tiers.

My shortlist

• simuland.ai — AI-native supply-chain simulation. This is probably the most directly aligned with what you described. Its Spenta platform is specifically positioned as an AI agent for supply-chain simulation and digital twins, including scenarios such as demand spikes, supplier failures, lead-time changes, DC/network redesign, and fulfillment strategies. Its general availability is planned for Q3 2026.
• anylogic.com — the simulation-engine specialist. Particularly strong if you want a rigorous, explainable model combining discrete-event, agent-based, and system-dynamics simulation. It has extensive supply-chain use cases and supports integrating AI/ML with simulations.
• simudyne.com — AI + massive-scale agent simulation. Very interesting if your twin needs to represent huge numbers of interacting entities—suppliers, products, customers, transportation nodes, etc. Simudyne says Amazon uses its technology to simulate the U.S. inbound supply chain involving hundreds of millions of products.
• siemens.com — industrial-scale Digital Twin platform. Especially compelling if your supply chain twin extends into factories, warehouses, material handling, production and physical assets. Siemens combines physics-based simulation, operational data and AI; its 2026 work with PepsiCo includes high-fidelity warehouse/facility twins and supply-chain simulation.
• kpmg.com — implementation/strategy rather than a simulation engine. Worth considering if you need help turning the twin into an enterprise decision-making capability—governance, workflows, scenario planning and integration with planning/operations.
• saviantconsulting.com — custom AI/Digital Twin implementation. They specifically advertise AI-enabled twins for warehousing, supply chain and manufacturing, so they're worth investigating if you want a partner to build rather than simply license a platform.

If I were building your architecture

I'd separate the problem into three layers:

1. Digital Twin / simulation engine AnyLogic, Simudyne, Siemens, or potentially Simuland.

2. AI reasoning/optimization layer AI agents that formulate scenarios, choose experiments, optimize inventory/network decisions, and explain results.

3. Enterprise data layer ERP + WMS + TMS + OMS + supplier/customer data + transportation + external signals, continuously feeding the twin.

The key is not to build an LLM that pretends to simulate the supply chain. The strongest architecture uses AI to interrogate and operate a calibrated simulation, while the simulation engine handles the actual dynamics and constraints. AnyLogic explicitly supports this AI/simulation approach, while Simudyne is taking it toward AI reasoning models trained against simulation environments.

For your specific project, my first three conversations would be Simuland, AnyLogic, and Simudyne. Siemens moves into the top three if your twin includes detailed manufacturing/warehouse physical operations.

If you tell me your industry, approximate supply-chain complexity (e.g. number of SKUs/sites/suppliers), and whether you want to build vs. buy, I can narrow this to 3–5 vendors and compare them on AI capabilities, simulation fidelity, integration, implementation time, and likely cost.

Yes. If by “simulation AI” you mean AI combined with a dynamic simulation model so you can ask “what happens if demand jumps 30%, a supplier fails, a port closes, or we add a DC?”, there are several specialists worth looking at.

My shortlist

• AnyLogic — probably one of the strongest foundations for a supply-chain digital twin. It supports discrete-event, agent-based, and system-dynamics simulation and has specific supply-chain/network capabilities through anyLogistix. It has demonstrated digital twins that combine simulation with ML and live supply-chain data.
• Cosmo Tech — particularly interesting if your goal is an AI-native, enterprise-level supply-chain twin rather than simply a simulation model. Their platform explicitly combines AI simulation with system-level digital twins and supply-chain scenario planning.
• Siemens — a strong choice if the twin extends from supply chain into factories, warehouses, machines and physical operations. Siemens combines physics-based simulation, operational data and AI; its 2026 KION partnership specifically applies AI, simulation and digital twins to supply-chain/warehouse operations.
• Miebach Consulting — very relevant if you need supply-chain expertise + simulation implementation, rather than buying a software platform and building everything yourself. They specifically position their work around living supply-chain digital twins and scenario analysis.
• Saviant Consulting — builds AI-enabled digital twins for warehousing, supply chain and manufacturing, including real-time data, future-event simulation and optimization.
• Simuland — an interesting newer entrant. Its focus is explicitly AI-native simulation for operational decisions, with supply chain, logistics and operations as core applications. Its team is currently delivering engagements using its Spenta simulation engine.

One important distinction

I would separate the market into three layers:

For example, Kinaxis is very strong in AI-powered supply-chain planning and orchestration, but I wouldn't put it in exactly the same category as AnyLogic/Cosmo Tech when the primary requirement is a dynamic simulation of the physical supply-chain system.

If this is a serious enterprise Digital Twin

I'd start with Cosmo Tech + AnyLogic + Miebach as three very different approaches and make them demonstrate the same use case.

The key test shouldn't be “show us your AI.” It should be:

Give us our network model, connect our actual data, introduce a disruption, and show us how the twin predicts the downstream effects and lets us evaluate alternative interventions. That's where you find out whether you're buying a genuine decision-making simulation or just another supply-chain analytics/control-tower product.

If you tell me your industry, approximate supply-chain complexity (plants/DCs/suppliers), and what you want the twin to optimize, I can narrow this to 5–7 firms that are genuinely appropriate and compare them by simulation technology, AI, implementation model, likely cost, and enterprise maturity.

Creating a digital twin for a supply chain requires combining network optimization, real-time data ingestion , and advanced simulation AI (such as discrete-event, system dynamics, or agent-based modeling). [1][2]

The market features several key players specializing in supply chain simulation, digital twins, and AI-driven scenario planning:[1][2][3]

• AnyLogic is a dominant force in multi-method simulation. It supports discrete-event, system dynamics, and agent-based modeling in a single platform, allowing you to model complex warehouse flows, transport networks, and inventory behavior. They also provide cloud scaling and integration capabilities with platforms like NVIDIA Omniverse for 3D/spatial simulations. [1][2][3]
• Coupa (via its LLamasoft heritage) is a gold standard for supply chain design, optimization, and digital twin scenario modeling. Coupa allows organizations to build an interactive digital model of their end-to-end network to test disruptions, evaluate inventory policies, and optimize multi-echelon flows. [1][2]
• Cosmo Tech specializes in Augmented Intelligence and Digital Twin simulation for complex industrial and supply chain systems. They focus heavily on prescriptive AI and "what-if" simulation engines that help decision-makers understand systemic risk and find optimal paths during disruptions.[1][2][3][4][5]
• Kinaxis integrates concurrent planning with advanced orchestration. Partnering heavily with accelerators like NVIDIA (utilizing GPU-accelerated optimization frameworks), Kinaxis Maestro provides concurrent, real-time insights and simulation capabilities across multi-tier enterprise networks. [1][2][3]
• IBM leverages AI-driven digital twin platforms embedded within its supply chain and asset management suites. IBM’s solutions ingest real-time IoT and POS data to balance inventory, forecast demand, and simulate network behavior under stress. [1][2][3][4][5]
• Oracle integrates IoT and digital twin simulation tools directly into its SCM and ERP cloud ecosystems, enabling businesses to mirror physical assets, run virtual simulations, and align data streams directly with accounting and logistics.[1][2][3][4][5]

To help narrow down which provider fits your architecture best, tell me:

• Are you focusing more on factory/warehouse floor simulation or macro-level end-to-end network routing?
• Do you need a platform that natively integrates with an existing ERP like SAP or Oracle?

Building a supply chain digital twin with simulation AI requires platforms that blend heavy data ingestion, multi-method simulation (discrete event, agent-based), and machine learning.

Several top companies and platforms specialize specifically in simulation AI and digital twins for supply chains:

Specialized Simulation Software Vendors

• AnyLogic : They are the industry gold standard for end-to-end supply chain simulation. Their specialized tool, anyLogistix , blends analytical optimization with deep simulation AI. It excels at analyzing risk, capacity bottlenecks, and complex "what-if" scenarios across multi-echelon networks.
• Cosmo Tech : This company focuses entirely on Enterprise Digital Twins powered by simulation AI. They specialize in vulnerable, complex supply chains, allowing companies to simulate cascading risks, financial impacts, and mitigation strategies before they happen.
• CreateASoft (Simcad DT) : They offer a patented on-the-fly digital twin modeler (Digital Twin Studio® ). It integrates live warehouse management (WMS) and transit data to dynamically simulate routing and logistics bottlenecks.

Industrial Tech & Enterprise Software Giants

• Siemens (Digital Industries Software) : Siemens provides industrial-grade digital twin tools like Tecnomatix and Plant Simulation . They specialize in mirroring physical warehouse layouts, automated conveyor systems, and micro-level material flows.
• Coupa (formerly LLamasoft) : A massive player in supply chain design. Their Supply Chain Guru platform allows organizations to build detailed digital twins to model inventory positioning, network footprint changes, and transportation dynamics.
• SAS Institute : Utilizing SAS Simulation Studio alongside their advanced AI and machine learning tools, SAS helps organizations build highly customized digital twins to manage complex multivariable operations.

Big Tech Infrastructure & AI Tooling

• NVIDIA (Omniverse & Isaac) : If you need to build a heavy, physics-based 3D digital twin of fulfillment centers, NVIDIA specializes in providing the underlying AI infrastructure. Their tech is heavily used to train autonomous robots and simulate macro-level warehouse optimization.

To help point you toward the right vendor, what is the primary goal of your digital twin (e.g., testing warehouse automation, optimizing global freight routes, or predicting component shortages)? Sharing your current ERP/data architecture would also help narrow it down

Yes. For a supply-chain Digital Twin specifically focused on simulation + AI, I’d shortlist these vendors:

• AnyLogic — Probably the strongest starting point if your core requirement is dynamic supply-chain simulation. It supports discrete-event, agent-based, and system-dynamics modeling, and can incorporate real-time data into digital twins. Its supply-chain platform also combines simulation with optimization.
• Cosmo Tech — Particularly interesting for AI-driven scenario analysis and supply-chain risk. Its platform creates dynamic digital twins to simulate how disruptions and decisions affect the overall network.
• Siemens — A strong choice if your twin extends from manufacturing → warehouse → logistics → supply chain. Siemens explicitly offers logistics digital twins for scenario experimentation and simulation, and its 2026 work with NVIDIA and PepsiCo is pushing AI + digital twins across industrial and supply-chain operations.
• Simio — Worth looking at when you want discrete-event simulation with embedded AI/ML optimization. Simio has native neural-network capabilities and can generate synthetic training data from simulation models.
• Simuland — A newer, particularly interesting entrant. Its Spenta product is positioned as an AI-native simulation/digital-twin agent specifically for supply chain, with the AI reasoning layer constructing simulation logic. As of August 2026, it's moving from early access toward broader availability.

How I'd think about the landscape

If you're building a serious enterprise twin, there are really three different capabilities to distinguish:

The distinction is important: Kinaxis, for example, is excellent at AI-powered supply-chain orchestration and planning, but that's different from having a simulation engine that lets you run thousands of hypothetical network states.

My first three calls would be AnyLogic, Cosmo Tech, and Siemens, depending on your architecture. I'd add Simuland if you're specifically interested in the emerging AI agent that builds and operates the simulation model rather than conventional simulation software.

If you tell me your industry, supply-chain scope (supplier → factory → DC → customer), and what you want the twin to predict/optimize, I can narrow this to 3–5 vendors and compare them on simulation fidelity, AI capabilities, APIs/data integration, implementation effort, and likely cost.