The platform will first be deployed at Dangote Petroleum Refinery in Lekki, Nigeria, one of the world’s largest single-train refineries.
Honeywell Technologies has launched Forge Plant Cognition, an industrial AI platform developed using NVIDIA AI technology and Dell Technologies’ AI infrastructure. The platform combines industrial process knowledge, operating data and AI tools to help refinery, petrochemical and liquefied natural gas (LNG) operators identify process constraints, anticipate problems and improve plant operations.
The platform will first be deployed at Dangote Petroleum Refinery in Lekki, Nigeria, one of the world’s largest single-train refineries. The project aims to increase throughput and support the work of local engineering and operations teams.
Based on preliminary data, Honeywell estimates that AI-enabled optimisation could deliver operational benefits exceeding US$0.29 per barrel at the refinery. At full production capacity, this could amount to more than US$60 million in gross annual benefits at the facility. These figures are company estimates and have not been presented as independently verified results.
Forge Plant Cognition combines real-time operating data with engineering knowledge and process models. It is intended to help operators identify production bottlenecks, detect developing equipment and process problems, and assess possible corrective actions.
The platform focuses on three areas: identifying constraints that affect throughput and product yield; detecting emerging problems that could lead to production losses or unplanned downtime; and making specialised process knowledge available to engineers and operators to support troubleshooting and operational decisions.
The system builds on Honeywell’s Forge Cognition platform and incorporates process expertise from Honeywell UOP. Its AI infrastructure uses Dell AI Factory with NVIDIA, supported by Dell Services. NVIDIA technologies, including Nemotron models, Omniverse libraries, Cosmos models and PhysicsNeMo, support data interpretation, digital-twin simulation, visual anomaly detection and physics-based modelling.
Deployment at Dangote refinery
At Dangote Petroleum Refinery, the platform will combine plant operating data with Honeywell’s process expertise to help identify operating constraints, investigate problems and inform decisions.
David Bird, CEO and Managing Director of Dangote Petroleum Refinery and Petrochemicals FZE, said the company selected the platform to support process optimisation and operational decision-making.
The project also aims to develop the skills of Nigerian engineers and operators by making specialised process knowledge available through AI-powered tools. Dangote has stated a goal of building a workforce that is 70% Nigerian, reducing reliance on external expertise over time.
The launch builds on a collaboration between the two companies. In November 2025, Honeywell announced that it would help Dangote increase the refinery’s capacity from 650,000 to 1.4 million barrels per day within three years, through the optimisation of existing assets and other improvements.
The performance of the new platform will depend on its implementation and the operational results achieved at the refinery.
