AspenTech

Learning Objectives

• Understand what AspenTech means by Industrial AI and how a hybrid model differs from a pure machine-learning model
• Evaluate the main AspenTech products that carry AI features, from predictive maintenance to the AVA platform
• Recognize how AspenTech is bought and what its Emerson ownership means for buyers

What Is AspenTech?

AspenTech (Aspen Technology) makes software that engineers and operators use to design, run and maintain process plants: refineries, chemical plants, gas processing facilities, power grids and similar capital-intensive operations. Its best-known products are the process simulators Aspen Plus and Aspen HYSYS, and its portfolio also covers predictive maintenance, advanced process control, production planning and scheduling, industrial data management, subsurface engineering and electric grid management.

The company grew out of the ASPEN Project, joint research between the Massachusetts Institute of Technology and the US Department of Energy prompted by the 1970s energy crisis, and is headquartered in Bedford, Massachusetts. AspenTech is now a wholly owned subsidiary of Emerson. Emerson took a 55 percent stake in May 2022, when it combined its own industrial software businesses with AspenTech, and on March 12, 2025 it completed the purchase of the shares it did not already own. AspenTech's stock stopped trading at that point, and the business now sits inside Emerson's Control Systems and Software segment.

AspenTech calls its approach Industrial AI: the company says it combines the speed of AI algorithms with the constraints of real-world engineering expertise. In practice, most of what it sells is physics-based simulation and mathematical optimization, which is not machine learning. The AI features are added on top of those models.

Pricing

AspenTech does not publish a price list. Its software is licensed to enterprises through AspenTech and Emerson sales channels, with terms based on the products and scale involved.

For a hands-on look before talking to sales, Emerson points to AspenTech.ai, a web-based interactive experience where industry professionals can try AVA's AI-assisted decision-support use cases.

Core Features

Hybrid Models in the Simulators

Aspen Hybrid Models, introduced in May 2021, embed machine learning inside Aspen Plus and Aspen HYSYS. The idea is to keep the chemistry and physics of a first-principles model while using plant data and machine learning to keep the model matched to how a real unit actually behaves. AspenTech groups them into three kinds: AI-driven, first-principles driven and reduced-order models, the last of which run fast enough to use inside planning and real-time optimization.

Predictive Maintenance With Aspen Mtell

Aspen Mtell, which AspenTech acquired in October 2016, uses machine learning on real-time and historical equipment data to learn each asset's normal behavior and warn of failures well before a breakdown. A January 2026 release added industry and asset templates for faster rollout, AI that groups and prioritizes alerts and prescribes corrective actions, and direct links to Emerson's AMS vibration monitoring software.

Aspen Virtual Advisor

Aspen Virtual Advisor (AVA) began as advisors built into individual products. AVA for Aspen DMC3 gives advanced process control engineers ranked recommendations grounded in established operating practice, and AVA for Aspen Unified PIMS does the same for refinery planning models. Both were expanded in version 15 of the software in May 2025.

The AspenTech AVA Platform

In May 2026 Emerson launched AspenTech AVA as a platform of its own. Emerson describes it as agentic and domain-aware: it pairs large language models with the company's first-principles models and embeds AI-assisted recommendations directly in operating workflows. It launched with four operational optimization and decision-support advisors.

Generative AI for Plant Layout

Version 15 also added generative AI to Aspen OptiPlant 3D Layout, which produces multiple digital facility layout options that weigh business, safety, engineering and construction requirements, for both new plants and existing sites.

Industrial Data Fabric

AI in a plant depends on usable data, and AspenTech Inmation is the company's data layer. AspenTech acquired inmation, a German industrial data company based in Cologne, in October 2022. In May 2026 Emerson rebuilt the Inmation operational technology data fabric on a distributed, node-based architecture meant to scale from one plant to a global enterprise and serve as the foundation for its AI tools.

Strengths

• Deep domain models — the AI sits on top of simulation and control software that process engineers already rely on, rather than replacing it
• Hybrid approach — combining physics with plant data keeps models grounded in known chemistry while tracking real equipment
• Broad coverage — one vendor spans design, control, planning, maintenance and data management across the plant lifecycle
• Emerson backing — ownership by a large automation company ties the software to Emerson's control systems and instrumentation
• Tangible AI features — predictive maintenance and model calibration are concrete, well-defined uses rather than general-purpose chat

Limitations and Considerations

• No public pricing or trial — access runs through enterprise sales, so it is out of reach for individuals and small teams
• Much of it is not machine learning — the Industrial AI label covers products whose core is physics-based simulation and optimization
• The agentic platform is new — AVA launched as a platform in May 2026 with four advisors and no customer results published alongside it
• Data dependency — hybrid models and predictive maintenance are only as good as the plant data behind them
• Ecosystem pull — deeper integration with Emerson hardware and software can make it a stronger fit for Emerson sites than for plants standardized on another automation vendor
• Specialist skills — the tools assume trained process, control and reliability engineers

Best Use Cases

Key Takeaways

• AspenTech builds industrial software for process plants, anchored by the Aspen Plus and Aspen HYSYS simulators
• Its Industrial AI approach layers machine learning and generative AI on top of physics-based models, and much of the portfolio is not machine learning at its core
• Aspen Mtell predictive maintenance and Aspen Hybrid Models are its most established AI features, while the agentic AspenTech AVA platform launched in May 2026
• AspenTech has been a wholly owned subsidiary of Emerson since March 12, 2025, and its shares no longer trade
• It is enterprise software sold through sales, with no public pricing or self-serve access