Shaping AI to deliver ‘green’ operating theatres

Artificial intelligence can be a powerful enabler of sustainability in healthcare, aligning environmental and financial benefits

Operating theatres are among the most resource-intensive environments in modern healthcare. They consume up to six times more energy per metre square than the rest of the hospital, generate as much as one-third of total hospital waste, and contribute disproportionately to the healthcare sector’s environmental burden.

In Ireland, as across all high-income healthcare systems, the imperative to reduce this carbon footprint is a matter of urgency. The emergence of artificial intelligence (AI) as a practical, deployable technology, which can facilitate the delivery of a ‘greener’ surgical service, is worthy of evaluation.

AI and the operating theatre

The Department of Health’s recent publication AI for Care – The Artificial Intelligence Strategy for Healthcare in Ireland 2026–2030 outlines a vision of this technology as a tool to be deliberately and ethically harnessed, aligning with core values of the Irish healthcare system and not as a disruptive external force.

This framing of AI places the multidisciplinary healthcare team as active architects of its implementation.

For the surgical community in Ireland, this offers a novel opportunity to utilise innovative technology to drive down the carbon footprint of the operating theatre.

The modern operating theatre was described as a ‘spaceship’ by UK Consultant Anaesthetist Dr Annie Hunningher at the Global Innovation and New Technology (GIANT) Health Event 2025. This comparison captures the pace of technological integration in perioperative environments. Digital pre-operative planning, continuous physiological monitoring, and AI-enabled thematic analysis of incident data are already reshaping how surgical teams understand and improve their performance. The trajectory toward 2030, which is the final year of the new strategy, takes this further.

In the context of sustainability, AI offers specific and measurable contributions across four domains. The first of these is in relation to energy management. AI-integrated systems can dynamically adjust heating, ventilation, and air conditioning – which accounts for between 90 per cent and 99 per cent of an operating room’s energy use – based on real-time occupancy data, potentially cutting consumption by up to 30 per cent. In a carbon footprint audit conducted in three academic hospitals, an HVAC setback system reduced energy consumption by 50 per cent overnight and at weekends without any impact on patient care. Scaling this kind of intervention nationally, with AI optimising setback decisions in real time across multiple theatres and sites, represents a significant and achievable emissions reduction.

Second, AI-powered computer vision systems can classify and segregate waste at the point of generation. This matters because waste audits consistently show that up to 90 per cent of operating room waste is misclassified. The majority is currently treated as biohazardous when it is not, which results in unnecessary costs and the environmental harm of incineration or autoclaving. Educational interventions alone have been shown to improve correct segregation and reduce costs by 75 per cent. AI-assisted real-time classification could sustain and exceed these gains without ongoing dependence on behavioural change alone.

Third, AI-driven predictive scheduling platforms – including systems already in clinical deployment such as Opmed.ai – maximise theatre utilisation by reducing idle time, cancellations, and the wasteful preparation of consumables for cases that do not proceed. High cancellation rates were identified in the GIANT 2025 panel discussion as both a patient safety concern and a financial liability. Intelligent scheduling addresses both simultaneously.

Fourth, at the level of surgical planning, AI supports the adoption of reusable energy devices in procedures such as laparoscopic colectomy, where reusable alternatives have been shown to reduce carbon emissions by between 40 per cent and 66 per cent compared to disposable equivalents. AI-assisted surgical planning tools can embed reusability preferences into the default configuration of operative lists, making the sustainable choice also the default choice.

As Prof Aldo Faisal, Professor in AI and Neuroscience, Imperial College London, UK, says: “AI must be designed to work together with people to get the best of both – not to do everything on its own.”

The Irish context

The HSE is taking climate action and outlines a strategic approach to the delivery of greener models of healthcare by 2050. What is now required is to align the Climate Action Strategy 2023–2050 with the AI for Care document to create a sustainable health system. An Irish perioperative sustainability dashboard, drawing on AI-generated data from theatre management systems, energy monitoring platforms, and waste classification tools, would provide the kind of actionable, transparent performance data that frontline teams currently lack.

Ms Stella Vig, Deputy National Medical Director for Secondary Care, NHS England, speaking at GIANT 2025, identified the absence of this data as one of the central barriers to quality improvement in operating theatres. Teams cannot improve what they cannot see.

The governance architecture for such a programme already exists in outline. The HSE National Clinical Programmes in Surgery and the HSE National Health Sustainability Office are best placed to translate the integration of AI into climate action policy at an operational level.

Navigating the tensions

It would be intellectually dishonest to present AI as an uncomplicated solution. There are genuine tensions to be navigated. The data infrastructure supporting AI is itself energy-intensive. To be credibly ‘green’, AI deployment in healthcare must prioritise efficient model architectures and renewable energy sourcing for the underlying infrastructure. The concept of ‘tiny AI’ – lean, purpose-built systems designed for specific clinical tasks rather than general-purpose large language models – should be considered as an option. Fit-for-purpose AI carries a smaller environmental footprint than general-purpose AI deployed without consideration of efficiency.

To be credibly ‘green’, AI deployment in healthcare must prioritise efficient model architectures and renewable energy sourcing for the underlying infrastructure.

Data security and integration present a further challenge. Irish hospitals, like their counterparts internationally, operate across fragmented legacy systems. General Data Protection Regulation compliance, data governance, and the technical complexity of integrating AI tools with existing theatre management and electronic health record systems require sustained investment and careful institutional planning. These are solvable problems, but they are not trivial ones. They require resource commitment at the systems level, not merely at the level of individual enthusiasts.

Finally, there is the human dimension. Clinicians’ reluctance to engage with new technology is not irrational. It frequently reflects, as Dr Hunningher observed at GIANT 2025, a lack of meaningful performance data and a sense that systems are imposed rather than co-designed. The lessons from successful ‘green’ operating theatre programmes are that implementation requires cultural embedding alongside technical deployment. 

How we shape AI

The operating room of 2030 will be shaped by the decisions we make today. The surgical community can leverage ‘lean AI’ to decarbonise the operating theatre. AI-driven solutions – which optimise energy usage, reduce waste, and improve resource management – can significantly reduce the carbon footprint of operating theatres while maintaining high standards of patient care.

References available on request