Honeywell Launches Forge Plant Cognition to Bring AI-Powered Optimization to Dangote Refinery

Industrial facilities are increasingly exploring artificial intelligence to improve efficiency, reduce unplanned downtime and make better use of existing equipment. Honeywell Technologies is bringing this approach to the energy sector with the launch of Forge Plant Cognition, an industrial AI platform being introduced at the Dangote Petroleum Refinery in Lekki, Nigeria.

Announced on October 26, 2026, the platform combines Honeywell's industrial process expertise with AI and accelerated computing capabilities from NVIDIA and AI infrastructure from Dell Technologies. The initiative aims to help refinery operators identify operational constraints, improve production performance and support faster decision-making.

Honeywell estimates that the platform could deliver more than $60 million in annual gross operational benefits at the Lekki refinery when operating at full capacity. This is a preliminary estimate, not a guaranteed financial outcome, and it excludes implementation costs.

What Is Honeywell Forge Plant Cognition?

Forge Plant Cognition is an industrial AI solution built on Honeywell's Forge Cognition platform. It combines operational data, engineering knowledge, industrial automation and physics-based models to help teams understand how a complex facility is performing and where improvements may be possible.

Rather than focusing only on displaying equipment data or generating recommendations, the platform is designed to connect operational insights with actions that can improve plant performance.

For example, refinery operators must manage interconnected processes where changes in one area can affect production elsewhere. Identifying a bottleneck requires an understanding of equipment conditions, process behaviour and operating constraints.

Honeywell says Forge Plant Cognition is designed to combine these factors so engineers can investigate issues, evaluate potential improvements and make more informed operational decisions.

The initial deployment at Dangote's refinery will integrate plant operating data with Honeywell's process expertise and AI-powered optimization capabilities.

Three Main Ways Forge Plant Cognition Aims to Improve Operations

Honeywell identifies three principal areas where the platform is designed to create value.

1. Identify hidden production value

Industrial facilities can lose potential output when equipment or interconnected processes limit overall production. Forge Plant Cognition is designed to help identify these bottlenecks and support optimization of throughput, product yield and product value.

For a refinery, better coordination between processes could help operators understand where operational adjustments may improve performance. The actual results will depend on the facility's configuration, operating conditions and other factors.

2. Protect production and reduce downtime

Unexpected equipment failures and process disturbances can interrupt production. The platform is designed to detect emerging equipment or process issues, recommend corrective actions and help minimize unplanned downtime.

This approach could give engineering teams additional information to investigate developing problems before they cause larger operational disruptions. However, Honeywell has not guaranteed that the platform will prevent failures or eliminate downtime.

3. Make industrial expertise more accessible

Refineries depend on specialized knowledge of equipment, chemical processes and operating conditions. This expertise can be difficult to distribute consistently across large teams and complex facilities.

Forge Plant Cognition is designed to embed industrial process knowledge into AI-powered workflows. According to Honeywell, this can help engineers and operators access expert guidance, solve problems faster and make more consistent decisions.

These capabilities are particularly relevant to the Dangote project, where developing local engineering and operational expertise is also a stated objective.

How NVIDIA and Dell Technologies Power the Platform

Forge Plant Cognition brings together several technologies rather than relying on a single AI model. Honeywell supplies industrial process knowledge and automation expertise, while NVIDIA technologies provide AI, simulation and accelerated computing capabilities. Dell Technologies supplies the supporting AI infrastructure, with Dell Services also supporting the platform.

The announcement identifies several specific components.

NVIDIA Nemotron

Open models power AI agents designed to understand and interpret operational data. This provides part of the platform's AI-driven analysis capability.

NVIDIA Omniverse

Omniverse libraries help create a digital twin of the plant and its processes. A digital twin is a virtual representation of a physical system that can support simulation, training and planning.

NVIDIA Cosmos and Metropolis VSS Blueprint

These technologies support vision AI agents designed to provide real-time alerts and detect anomalies.

NVIDIA PhysicsNeMo

The framework is used to build, train and fine-tune physics AI models at scale, supporting digital-twin-integrated plant optimization workflows.

Dell AI Factory with NVIDIA

Dell provides the AI infrastructure behind the platform, with Dell Services supporting the deployment.

The combination is intended to connect industrial knowledge with AI interpretation, physical-system simulation and operational data. The announcement does not disclose a complete deployment architecture, model configurations or detailed infrastructure specifications.

Dangote Refinery: The First Deployment

The Dangote Petroleum Refinery in Lekki, Nigeria, is the first announced deployment of Forge Plant Cognition.

The project is intended to integrate the refinery's operating data with Honeywell's process expertise and AI-powered optimization capabilities. The stated objectives include identifying production constraints, accelerating troubleshooting and supporting more informed operational decisions.

The project also connects industrial AI adoption with workforce development. Honeywell says the platform will help local engineers and operators access specialized process knowledge through AI-powered tools.

Dangote has stated a goal of developing a workforce that is 70% Nigerian as it expands, with the intention of reducing reliance on external expertise over time. The announcement presents the platform as one way to support that objective, rather than as a guarantee that the workforce target will be achieved.

What about the projected $60 million benefit?

Honeywell's preliminary estimate is based on potential operational benefits of more than $0.29 per barrel at the Lekki refinery. At full-capacity throughput, the company estimates that this could amount to more than $60 million in gross operational benefits annually for the facility.

These figures require important context:

• They are preliminary estimates, not verified realized savings.

• They apply specifically to the Lekki refinery.

• The estimate is gross of implementation costs, which are excluded.

• Actual outcomes depend on operating conditions, refinery configuration, crude oil mix, utilization, baseline performance, data quality, product margins and implementation.

Consequently, the $60 million figure should be understood as a projected opportunity under the stated assumptions, not a confirmed financial result.

What This Development Means for Industrial AI

Honeywell's launch highlights a broader industry shift toward using AI in physical environments, where decisions affect equipment, production processes and operational reliability.

Traditional industrial monitoring systems help teams observe equipment conditions and identify deviations. Industrial AI platforms aim to go further by combining operational data with specialized engineering knowledge, simulation and decision-support capabilities.

For enterprises, this approach could offer several potential benefits:

• Better operational visibility: Connect information from complex processes to help teams understand performance constraints.

• Faster troubleshooting: Help engineers investigate emerging problems using operational data and embedded process knowledge.

• Improved planning: Use digital twins and simulation to evaluate operating scenarios.

• Knowledge sharing: Make specialized engineering expertise more accessible to local teams.

These are potential benefits based on the platform's stated design and intended uses. Actual performance will depend on implementation, data quality, the operating environment and how teams use the system.

Industrial AI also brings practical considerations. Operators must evaluate how AI-generated recommendations fit existing engineering procedures, safety requirements and operational controls. The announcement does not disclose detailed information about safety validation, cybersecurity controls, human approval requirements or the extent of autonomous decision-making.

Conclusion

Honeywell's Forge Plant Cognition launch at the Dangote Petroleum Refinery demonstrates how industrial AI is being applied to complex energy operations. By combining industrial process expertise with NVIDIA AI and simulation technologies and Dell infrastructure, the platform aims to help refinery teams optimize production, identify emerging problems and strengthen engineering capabilities.

The projected annual gross operational benefit of more than $60 million at full capacity is notable, but it remains a preliminary estimate rather than a confirmed outcome.

The deployment will be relevant to industrial operators evaluating how AI, digital twins and physics-based models can support real-world operations. Its longer-term significance will depend on the operational and financial results achieved in practice.