3 minute read

Sean Lockie, Associate Director at Arup

Steel, cement and chemicals form the backbone of modern economies, but they also account for roughly 25% of global carbon emissions. These sectors are notoriously hard to abate as their processes rely on extreme heat and carbon intensive reactions. As climate targets tighten and carbon pricing escalates, the race to decarbonise these industries is no longer optional - it’s essential. Enter Artificial Intelligence (AI): a transformative force accelerating the shift to low carbon production.

Why these sectors are so challenging

Steel production alone contributes about 7% of global CO₂ emissions, primarily from blast furnaces and coking operations. Cement adds another 8%, driven by the calcination of limestone and fossil fuel combustion in kilns. Chemicals, including ammonia and plastics, emit vast amounts of CO₂ and nitrous oxide during synthesis processes. Traditional mitigation strategies such as alternative fuels, and electrification remain vital but cannot deliver the scale of change required without systemic efficiency gains. This is where AI and circular economy principles can accelerate progress.

AI as an optimisation engine

AI is already helping the industry make smarter decisions. In steelmaking, advanced algorithms analyse sensor data from blast furnaces, adjusting temperature, pressure and raw material mix to minimise energy use and emissions. Predictive maintenance reduces downtime and prevents energy waste, while computer vision monitors particle sizes for optimal combustion.

Similar gains are seen in cement, where AI-driven kiln optimisation and digital twins - a virtual representation of the kilns - fine tune fuel mix and heat distribution, cutting fuel-derived emissions by up to 5% per plant annually. Machine learning can also be used to optimise concrete mix designs, delivering lower carbon outcomes.

While AI excels at optimisation, designing fundamentally new processes, materials and business models still requires human creativity and interdisciplinary expertise. That’s where circular thinking becomes critical.

Circular economy: Beyond reduce, reuse, recycle

Decarbonisation isn’t just about cleaner production - it’s about rethinking the entire lifecycle. The “9 R’s” framework reminds us that redesign is as important as reuse. For steel and cement, this means designing for disassembly, enabling components to be recovered and repurposed rather than demolished. Today, dismantling is blunt and labour intensive, but tomorrow, robotics and machine learning could automate sorting and recovery. Materials passports embedded in building information models (BIM) and digital twins will allow precise tracking of what’s in our buildings, which is essential for urban mining and reuse.

Arup is already exploring circular approaches through urban mining pilots in Europe, mapping building technologies to identify concrete suitable for recycling. The consultancy is also driving standardisation of building elements like glass partitions and ceiling tiles, which can further unlock reuse potential. As part of one of its many research investment programmes, Arup is currently developing DeepEnergy, an AI-enabled framework which will help clients plan decarbonisation across entire portfolios, even when operational data is incomplete.

Additionally, industrial symbiosis - sharing waste heat between facilities - is another opportunity, already being trialled in eco-industrial parks such as Kalundbord in Denmark.

Mindset matters

AI is a powerful enabler, but it works best as part of the broader strategy. It can optimise processes, predict outcomes, and accelerate progress, but lasting impact depends on collaboration across industry, academia and policy. Success also requires openness to new materials, new design approaches and new data models, especially in emerging markets where emissions data is scarce. A prime example is the Global Whole Life Carbon Tracking (GWLCT) project, funded by the UK government Department for Energy Security and Net-Zero (DESNZ). The initiative is developing global carbon tracking methods and tools, providing data and insights to accelerate decarbonisation efforts, especially in developing economies, and highlighted opportunities for lower carbon construction through analysis across the world. Part of the method involves using earth observation and machine learning to track the changes in built environment areas.

Bottom line

Decarbonising steel, cement and chemicals demands both AI and circular thinking. AI helps us move faster and make better decisions, helping these carbon-rich industries cut emissions at scale. But real transformation comes though combining these digital capabilities with redesign, reuse and systemic change. Together, these approaches can turn ambition into action and shape a built environment that supports a net zero future.

You May Also Enjoy