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

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).

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

• 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.
• 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.
• 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.
• 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.
• 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.
• 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.

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.

When building a digital twin for a supply chain that incorporates AI and advanced simulation, different providers excel depending on whether you need macro-level network design, factory/warehouse-level physics simulation, or enterprise software integration.

The market leaders and specialists in simulation AI and supply chain digital twins are organized below by their core strengths:

1. Dedicated Simulation & Network Modeling Engines

• AnyLogic / anyLogistix (ALX): The gold standard for complex, multi-method simulation (combining discrete-event, agent-based, and system dynamics modeling). Their platform anyLogistix is purpose-built for supply chain design and risk-stress testing.
• Simio: Integrates intelligent object-based 3D simulation with embedded machine learning and AI optimization to test workflow, capacity, and logistics scheduling.

2. Industrial Giants & Physics-Based AI

• Siemens: A dominant force across industrial and logistics operations. Through platforms like Siemens Xcelerator and their heavy investment in enterprise simulation (including their acquisition of Altair), they excel at connecting operational tech, facility process simulation, and supply chain flows.
• NVIDIA (Omniverse & Isaac Sim): The pioneer of physical AI and real-time ray-traced simulation. NVIDIA collaborates with major integrators (like Accenture) to build industrial-scale omniverse digital twins for automated warehouses, robotics fleet management, and complex manufacturing transport networks.
• Ansys: Known as the pioneer for engineering digital twins, using heavy physics-based simulation models to predict node, asset, and system performance over time.

3. Enterprise SCM & Control Towers

• Coupa (Supply Chain Design and Planning): Excellent for end-to-end network optimization, continuous design, and financial alignment, allowing you to model multi-echelon inventory and test trade-offs.
• Dassault Systèmes (3DEXPERIENCE / DELMIA): Powerful for synchronized planning, scheduling, and 3D digital manufacturing/logistics ecosystems.
• ToolsGroup: Focuses heavily on AI-driven simulation and inventory intelligence to continuously mirror and respond to erratic market disruptions.
• SAP / IBM: Best if your digital twin needs to live directly inside an enterprise resource planning (ERP) backbone, ingesting live transactional data streams from WMS, TMS, and ERP suites to feed machine learning workflows.

To help narrow down which partner or software stack fits best, could you share:

• Are you focusing more on macro network design (sourcing, warehouse locations, transport lanes) or micro-operations (warehouse automation, shop-floor logistics, robotics)?
• Do you plan to build custom on a platform like NVIDIA Omniverse/AnyLogic, or deploy an out-of-the-box enterprise SCM suite?

Yes. If your goal is a Digital Twin of the end-to-end supply chain—suppliers → plants → inventory → transportation → DCs → customers—with AI running what-if scenarios, forecasting, optimization, and disruption response, I’d look at these specialists:

My recommendation

I'd separate the problem into two layers:

1. Simulation engine — "What will happen?"
Use something like AnyLogic if your core requirement is modeling the stochastic behavior of the supply chain: lead-time variability, queues, inventory policies, transportation delays, capacity constraints, supplier failures, etc. Its published case work is unusually close to what you're describing.

2. AI decision layer — "What should we do?"
Put an AI/optimization layer over the simulation that can ask thousands of scenarios automatically:

"What happens if Taiwan supply falls 30%?"
"Should we reroute inventory from DC A to DC B?"
"Where should we add capacity?"
"Which suppliers should we dual-source?"
"What's the optimal inventory positioning for a 95% service level?"
"What happens if transportation costs rise 20%?"

For a physical/warehouse/manufacturing-heavy twin, I'd seriously investigate the Siemens + NVIDIA combination. Their current architecture is remarkably aligned with this vision: Siemens provides industrial data, engineering and simulation expertise while NVIDIA supplies accelerated simulation/AI infrastructure. PepsiCo is already using the combination to create high-fidelity twins of plants and warehouses and simulate end-to-end supply-chain operations.

One important distinction

Don't let vendors sell you a "digital twin" that's really just a dashboard.

For what you're describing, I'd insist that the platform can:

Real-world data → calibrated simulation → AI-generated scenarios → optimization → recommended action → actual outcome → twin recalibration

That's the closed loop that turns a digital twin into a decision-making system, rather than a visualization.

If you tell me your industry, approximate supply-chain scope (e.g. 50 suppliers / 10 plants / 20 DCs / global transportation), and what decisions you want the twin to optimize, I can narrow this to 3–5 vendors/partners and give you a recommended architecture and RFP criteria.

When building a digital twin for a supply chain that incorporates artificial intelligence and predictive simulation, specialization spans across core simulation engines, enterprise AI platforms, and heavy industrial/logistics digital twin developers.

The primary market leaders and specialists in simulation AI for supply chains are categorized below by their core strengths:

• Core Simulation & Network Modeling Engines
  • AnyLogic (anyLogistix): Renowned for combining discrete-event, agent-based, and system dynamics simulation. Their anyLogistix platform is built specifically for end-to-end supply chain design, stress-testing, and risk-free optimization.
  • Simio: Specializes in discrete-event simulation and intelligent digital twins that model complex operational flows, material handling, and warehouse routing with high fidelity.
• Enterprise AI & Supply Chain Platforms
  • C3.ai: Provides the C3 AI Supply Chain Digital Twin framework, which layers machine learning and intelligent AI agents over high-level data models to automate demand planning, continuous "what-if" scenario testing, and risk forecasting.
  • Kinaxis (Maestro AI): Known for concurrent planning and orchestration, using advanced graph analytics and explainable AI to model end-to-end multi-echelon supply chain networks and audit decisions before execution.
  • Coupa: Offers powerful Supply Chain Design and Planning tools that use digital twin technology to map nodes, lanes, and financial variables to optimize inventory placement and logistics paths.
• Industrial Infrastructure & Visualization Giants
  • NVIDIA (Omniverse): Provides the foundational, physically accurate, and real-time AI simulation platform utilized by major enterprises and software vendors to bring 3D data-driven supply chain and factory replicas to life.
  • Siemens: A dominant force in industrial manufacturing twins through its Xcelerator portfolio, frequently partnering with heavy tech leaders like NVIDIA to implement physical AI and deep process simulation across factory and logistics ecosystems.
  • Dassault Systèmes: Leverages its 3DEXPERIENCE platform to build multi-scale virtual twins covering everything from manufacturing plant lines to the broader value chain.
• Consulting & Custom Integrators
  • BCG X: The tech build-up arm of Boston Consulting Group, which specializes in deploying customized Value Chain Digital Twins that ingest external data (port congestion, macro indices) alongside internal metrics to run complex probabilistic forecasts and simulations.
  • SimWell: A specialized consultancy and development firm focusing entirely on simulation-based digital twins for logistics, warehousing, and complex supply chain balancing.

To help narrow down which provider fits your architecture, could you share:

• Are you focusing more on macro-level network routing/inventory or micro-level warehouse/factory floor simulation?
• What data stack or cloud ecosystem (Azure, AWS, SAP, custom) are you planning to build on top of?

Yes. For a supply-chain Digital Twin where simulation + AI are central, I’d shortlist these players:

My recommendation

If you're building a true end-to-end supply-chain twin, rather than simply a visualization/control tower, I'd investigate AnyLogic + Cosmo Tech + Palantir first.

• AnyLogic → best if the heart of the project is simulation science: suppliers → plants → inventory → DCs → transportation → customers.
• Cosmo Tech → best if the key question is "What happens if X goes wrong or we make Y decision?" across a complex network.
• Palantir → best if you want to turn the twin into an AI-powered operational decision platform, not just a simulation model.
• Siemens/NVIDIA → add this when physical facilities, warehouses, robotics or factory-level physics matter. Siemens is explicitly positioning its Digital Twin technology around physics-based simulation combined with AI and real-time operational data.

One particularly relevant proof point: AnyLogic has already been used to build a supply-chain Digital Twin spanning vendors, distribution centers and final-mile sites, including forecasting order-to-delivery time and inventory allocation.

If you tell me your industry, scale (e.g. number of plants/DCs/SKUs), and what you want the twin to optimize, I can narrow this to 3–5 vendors/consultancies and compare their AI, simulation, data integration, implementation effort, and likely fit.

Building a digital twin for a supply chain requires combining network modeling, real-time data ingestion, and AI-driven simulation engines . Several enterprise technology leaders, simulation specialists, and consulting firms specialize in this space:

• Core Simulation & Digital Twin Platforms
  • AnyLogistix (ALX) (by AnyLogic) — Widely recognized for network optimization paired with heavy discrete-event and dynamic simulation for complex supply chains.
  • Siemens (Simcenter, Teamcenter) — Specializes in physical-to-digital process, plant, and logistics simulations, heavily expanding into industrial AI.
  • Dassault Systèmes (3DEXPERIENCE platform) — Offers robust virtual environments tailored to manufacturing ecosystems and end-to-end value chain modeling.
  • Ansys (Twin Builder) — A dominant force in physics-based and hybrid-analytics engineering simulations that translate into operational digital twins.
• AI Infrastructure & Hardware/Software Ecosystems
  • NVIDIA Omniverse — Provides the foundational graphics, cloud, and Physical AI simulation platform used by major logistics and industrial automation providers (like Siemens and Körber) to run physically accurate, large-scale virtual simulations.
  • Microsoft (Azure Digital Twins) — Supplies the underlying cloud infrastructure and IoT integration framework to build real-time spatial and operational twin environments.
• Supply Chain & Enterprise AI Specialists
  • Coupa / LLamasoft — Focuses extensively on network design, strategic planning, and risk-scenario modeling.
  • BCG X (Value Chain Digital Twin) — Offers advanced AI-based probabilistic forecasting and end-to-end network simulation as an advisory/tech-agnostic implementation partner.

If you'd like to narrow this down, please share:

• Are you focusing more on macro-network logistics (routing/sourcing) or micro-facility operations (warehousing/manufacturing floors)?
• Do you prefer a ready-made enterprise SaaS platform or a custom build using developer simulation engines?

Yes. If by “simulation AI” you mean a system that can take a digital twin of your supply chain and let AI agents run what-if scenarios, stress-test disruptions, optimize policies, and learn from simulated outcomes, there are a few specialists I would put on the shortlist.

My shortlist

The distinction I'd make

There are really three different markets hiding under “simulation AI”:

1. Supply-chain simulation → AnyLogic / Simuland
If your goal is:

“What happens to inventory, service levels, capacity, transportation and cost if supplier X fails, demand jumps 30%, or we change our replenishment policy?”

I'd start here. AnyLogic is particularly mature for this. Its digital-twin approach explicitly supports real-time data, scenario experimentation and AI/ML integration.

2. Industrial digital twin → Siemens + NVIDIA
If your twin includes:

suppliers → plants → warehouses → conveyors → robots → transportation → customers

and you eventually want AI agents to operate or optimize the physical system, I'd investigate the Siemens/NVIDIA stack. PepsiCo, for example, is using Siemens and NVIDIA to create high-fidelity twins of manufacturing and warehouse operations and simulate end-to-end supply-chain behavior.

3. AI-native “world model” → Simuland/Spenta
This is the most interesting emerging category to me. Rather than simply putting AI on top of a conventional simulation, the objective is to make simulation accessible to AI so that an agent can explore thousands/millions of possible operational futures and learn which decisions work. Simuland is explicitly positioning itself in this space.

What I would do for your project

For a supply-chain Digital Twin, I'd probably evaluate AnyLogic and Simuland first, then bring in Siemens/NVIDIA if the physical/warehouse/manufacturing layer is important.

The architecture I'd aim for is roughly:

Real supply-chain data
↓
Digital Twin / simulation model
↓
AI agents
↓
Millions of simulated scenarios
↓
Optimization / recommended decisions
↓
Human approval or autonomous execution
↓
Actual supply chain
↓
Feedback into twin

The key is that the twin shouldn't just be a visualization of the supply chain. It should become a decision laboratory where your AI can safely experiment with policies before you deploy them in the real network.

If you tell me your industry, approximate supply-chain complexity (e.g. number of plants/DCs/SKUs), and what you want the AI to optimize, I can narrow this to 5–7 specific vendors/teams and tell you who I'd actually invite to an RFP.