MORGANABLE INVESTMENT WATCH
Dangote Petroleum Refinery is adopting a new artificial intelligence platform to improve productivity, strengthen equipment reliability and help its teams make faster operating decisions.
Akure —
Dangote Petroleum Refinery is adopting a new artificial intelligence platform to improve productivity, strengthen equipment reliability and help its teams make faster operating decisions.
Honeywell Technologies announced the launch of Forge Plant Cognition on Monday, October 26, 2026. The system will be introduced first at the refinery in Lekki, Lagos, as the facility prepares for expansion.
The platform combines Honeywell’s industrial expertise with artificial intelligence and computing technology from NVIDIA and Dell Technologies.
It is designed to help refinery workers identify production bottlenecks, spot developing problems and choose practical responses. However, the financial benefits announced so far are estimates.
Honeywell estimates that the platform could create more than 29 cents in gross operational benefits for every barrel processed at the Lekki refinery.
At full-capacity throughput, that could amount to more than $60 million a year. The company stressed that the figure is preliminary, applies specifically to the refinery and is not guaranteed.
HOW THE AI PLATFORM COULD HELP
Forge Plant Cognition brings together live operating data, engineering knowledge and models of industrial processes.
It aims to help staff understand what is happening across the refinery and identify changes that could improve performance.
For example, it could help engineers find bottlenecks that limit production, improve the amount of useful products obtained from crude oil and increase their value.
It is also designed to identify early signs of equipment or process problems. By alerting staff and recommending corrective steps, the platform could help reduce unexpected shutdowns and costly delays.
The system will use several technologies developed by NVIDIA. Nemotron AI models will help software agents interpret operating data, while NVIDIA Omniverse tools will support a digital twin of the refinery and its processes.
A digital twin is a virtual model that allows teams to study operations, test scenarios and plan changes before applying them in the physical plant.
The platform will also use vision-based AI tools to identify unusual conditions and issue alerts. In addition, NVIDIA’s PhysicsNeMo framework supports the development of AI models that account for physical processes.
Dell Technologies will provide supporting AI infrastructure and services. Together, these tools are intended to connect operating information with engineering knowledge and help teams assess possible responses.
The technology is meant to support human decision-making, not replace experienced refinery workers.
Engineers and operators will still need to assess recommendations against plant conditions, safety requirements and operating procedures.
The platform’s value will depend partly on the quality of its data and how well it fits into existing systems.
EXPANSION AND EXPECTED BENEFITS
The launch comes as Dangote Refinery pursues a major capacity expansion. Honeywell is supporting plans to increase the refinery’s nameplate capacity from 650,000 barrels per day to 1.4 million barrels per day by 2028.
If completed, the expansion would raise the volume of crude the facility can process and increase the importance of reliable operations.
At a large refinery, even small improvements can matter when applied across many barrels. Better production flow, fewer equipment interruptions and improved product yields could help the company control costs and get more value from its assets.
Nevertheless, the projected annual benefit depends on the refinery reaching full-capacity throughput and the platform delivering the expected improvements.
Honeywell said the estimate excludes implementation costs. Therefore, the projected $60 million should not be described as confirmed profit or money already saved.
Actual results will depend on factors such as the refinery’s configuration, operating conditions, crude oil quality, utilisation, product margins and performance before the system is introduced. The companies have not published evidence of realised savings.
David Bird, chief executive and managing director of Dangote Petroleum Refinery and Petrochemicals, said the company chose the platform because it combines industrial knowledge with AI capabilities.
He said the investment was intended to improve performance, unlock more value from operations and support Nigerian talent.
Honeywell said the system would help local engineers and operators access specialised process knowledge through AI-supported tools.
This could help them solve problems faster, strengthen technical skills and make more consistent decisions.
Dangote aims to build a workforce that is 70 per cent Nigerian, and expects technology and training to support that goal.
For Nigeria, the deployment is another example of advanced digital technology entering a major industrial operation.
Refining depends on complex equipment and carefully managed processes, so problems can affect output, costs and reliability.
AI tools that help staff identify issues earlier may be useful, although their performance must be measured in real operating conditions.
The announcement also reflects a wider effort by industrial companies to use AI beyond office work and online services.
In refining, petrochemicals and liquefied natural gas, operators are exploring ways to apply data analysis and simulation to equipment and production systems.
Honeywell said Forge Plant Cognition is designed for refining, petrochemical and liquefied natural gas operators.
However, Dangote’s Lekki refinery is the first launch site identified in the announcement.
The announcement does not specify a timetable for full deployment or provide a breakdown of implementation costs.
Those details will matter when the refinery assesses the technology’s return and compares projected gains with operating results.
For now, the key development is the launch of the AI platform and its planned use at Dangote Refinery.
The next important measure will be its actual performance: whether it helps the plant increase throughput, prevent disruptions and improve product value.
Verified operating data will be needed to compare results with the company’s estimates.
If the platform delivers the expected gains, it could help Dangote manage its expanding operations while giving Nigerian engineers greater access to specialised technical guidance.
Still, the estimated $60 million in annual gross benefits remains a projection, not a confirmed result.
