Envision & KBR: AI optimized ammonia production

“Millions of tons” production capacity in the Gobi Desert

Click to learn more. Envision Energy has released further details about its AI architecture, with two “foundation models” working in tandem to optimize renewable energy inputs for ammonia production.

After reporting that its now-operational Chifeng renewable ammonia production plant in China was “fully AI-enabled” in 2025, Envision Energy has released further details about its AI architecture. Able to analyse critical incoming data streams like renewable energy generation, storage, and grid inputs, the “Dubhe Foundation Model” can help optimize operations for hydrogen & ammonia production. Dubhe works in tandem with “Tianji”, a foundation model analysing weather data, allowing the whole system to work “efficiently and reliably at scale”. Launched at Abu Dhabi Sustainability Week in January, Envision also signed a development and deployment partnership with Masdar.

Envision’s global-leading AI Power System intelligently schedules and balances the variability of wind and solar power in real-time. This ensures a constant, high-efficiency energy supply for the chemical synthesis process, significantly lowering the levelized cost of hydrogen.

Further details in Envision’s official press release, 26 Jan 2026

Subsequently at the World Economic Forum meeting in Davos, Switzerland, Envision unveiled plans to roll out the AI-enabled, renewable system to convert wind and solar energy in China into “millions of tons of green ammonia”:

Through our AI Power System, we are transforming the Gobi Desert into a ‘Green Oil Field,’ converting wind and solar resources into millions of tons of green ammonia,” Zhang stated. This approach reframes renewable energy from a volatile resource into a precise industrial tool, providing a scalable blueprint for global green industrialization.

Lei Zhang, Founder and CEO of Envision, in his organisation’s official press release, 26 Jan 2026

The Chifeng project features in the World Economic Forum’s (WEF) latest white paper, From Blueprint to Reality: A Stronger Business Case for Shared Energy Infrastructure, and has secured a “rigorous suite of high-standard certifications” (including ISCC EU, ISCC PLUS, and RFNBO compliance).

Combining with existing digital control platforms

At India Energy Week, Applied Computing and KBR announced the launch of INSITE 3.0, an AI-based platform designed to achieve low-emission, energy-efficient and reliable ammonia production. UK-based Applied Computing’s Orbital AI model will combine with KBR’s INSITE® digital platform, enabling users to “optimize day-to-day operations, reducing operational costs, increasing reliability and yield, and improving safety”.

Ammonia plants are some of the most complex industrial systems on the planet. KBR’s ammonia licensing relationships represent 50% of the world’s capacity and give us unprecedented reach to deploy INSITE 3.0 where it’s needed most. This solution is designed to unlock value that traditional systems simply cannot by processing 100% operational data in real time to deliver explainable, physics-grounded insights that operators can trust and act upon immediately.

Dan Jeavons, Applied Computing President, in his organisation’s official press release, 21 Jan 2026