Gigaton announced a $26 million funding round led by Plural with participation from 2150, Semapa Next, Planet A Ventures, AlbionVC, Cambridge Enterprise, and Clean Growth Fund. The London-based company is building machine learning systems to autonomously control large-scale industrial manufacturing processes, with a mission to reduce industrial carbon emissions by gigatonnes. Financial terms beyond the round size were not disclosed.
Gigaton’s approach sets it apart from most industrial AI companies. Rather than building chat interfaces, analysis tools, and recommendation agents, the company sits deep within the control stack of industrial facilities and operates plants autonomously through machine learning systems built by a team of machine learning scientists and chemical engineers. The company started by helping factories analyze and understand their processes, but through years of experience in control rooms, they determined that direct control of the underlying hardware was the only path to achieving their emissions reduction mission. The result is an AI that autonomously physically operates industrial facilities, such as a cement factory that operates at 1,500 degrees Celsius in a 60-meter-long, 15-story-high structure, in real time and autonomously from London.
Cement production alone accounts for about 8% of global CO2 emissions, and the approximately 5,000 cement plants around the world generate terabytes of operational data and spend billions of dollars annually on energy. Gigaton said this number represents only a fraction of the tens of thousands of industrial assets that share similar characteristics and are not being addressed by autonomous software. The company’s theory is that the physical world is already run by machines, and the software needed to control them autonomously hasn’t been modified. The new funding will support the continued deployment of Gigaton’s industrial AI control system across the world’s most emissions-intensive manufacturing processes.
Important quote:
“We’re not building chat interfaces or agents that will analyze and make recommendations. We’re building machine learning systems that sit deep in the control stack and actually run industrial plants. The world is already run by machines. These machines just haven’t been retrofitted with the software they need for autonomous control. We’re building it for the most important industry on the planet.”
Josh Vernon, CEO, Gigaton
