How AI can make the print industry more sustainable

October 14, 2024
Articles, Artificial Intelligence, Sustainability

AI can be a double-edged sword when it comes to environmental impact. Training complex AI models requires massive computing power, translating to huge amounts of power used by data centres. However, AI can also be a force for good, helping to optimise processes across the supply chain. How can the print industry use AI to streamline manufacturing processes, reduce waste, and lower their overall environmental impact?

It is a given that AI has the potential to drive economic growth and productivity. It will be able to help with decision-making across many areas within an organisation, such as optimising energy usage and overall carbon emissions. However, AI has its own environmental cost in how it meets its objectives through massive data centre resource usage. The challenge will be using it effectively to support an organisation’s transition to net zero. The ITC industry, including the print sector, is estimated to be responsible for 2–4% of global carbon emissions annually. The pulp and paper sector is also estimated to be responsible for around 2% of industry emissions. Many governments have set targets for meeting stringent carbon emissions targets by 2030, making this a pivotal year for organisations to meet their own and legally mandated targets. To help organisations achieve their own targets, print suppliers need to embrace AI as an important enabling technology.

AI can be a double-edged sword, so it is vital to understand the potential negative environmental impact. Training AI models and running queries within the data centres that provide AI infrastructure require immense computing power. This volume of data adds to storage and energy demands and increases carbon footprint. These factors drive up energy consumption and carbon emissions, particularly if fossil fuels are used for energy generation. Data centre cooling adds to the environmental impact in terms of water consumption. The rapid pace of AI development is also adding to the e-waste problem. Balancing AI’s positive and negative effects is one of the difficulties of adoption.

Tackling the carbon footprint of the print industry

The print industry has a challenging road to net zero. Unlike other technology products, the carbon footprint of a print or MFP device extends beyond the device to encompass paper and consumables such as ink and toner. These are all major contributors to environmental impact. Circular principles have long been embedded in the print industry. Today, many manufacturers have adopted a sustainable-by-design approach, with products designed for longevity and optimal energy efficiency and recycling programmes offered for both devices and consumables. Increasingly, suppliers are also offering refurbished and remanufactured products as part of their sustainability strategies.

Quocirca’s recent Sustainability Study reveals that 70% of IT decision-makers expect their print suppliers to not only demonstrate strong sustainability credentials across their business, products, and services but also help them measure their own environmental impact.

AI is emerging as a key enabler for sustainability in the print industry. Applying it across the product lifecycle can help optimise energy efficiency, uncover opportunities for digitisation to reduce paper usage, and enhance predictive maintenance models that minimise costly service engineer visits.

Applying AI to the print value chain

AI can be transformational across the print lifecycle, supporting energy efficiencies, driving optimisation on several fronts, and improving end-user experiences and sustainability credentials while assisting with legislative compliance.

• Energy efficiency. This is a major area with carbon emission reduction potential and a central theme across AI-enabled sustainability and optimisation applications. AI can help mitigate the impacts of climate change by optimising energy grids, selecting renewable energy, and developing smart solutions.
• Designing for lower emissions. AI for design enables energy efficiencies and waste-reducing support for recycling and repairs to be built in, helping meet energy efficiency and right-to-repair legislative requirements. Using AI to explore energy usage and operational and maintenance data from existing hardware enables designs to be adapted to improve energy efficiency and lower emissions of new builds.
• Service optimisation and predictive maintenance. Optimising IoT-connected multifunction printer (MPS) fleets through AI-driven insights is a high-value application area. Using data generated by IoT edge devices and sensors, AI algorithms can monitor the condition of MFPs in real time, detecting potential issues before they cause downtime and enabling proactive and predictive maintenance and service optimisation. This reduces the need for energy-consuming emergency repairs, lowers carbon emissions, and supports the efficient use of field service engineers, which in turn reduces emissions from fuel while also achieving lower device downtime and delivering faster service resolutions for customers.
• Demand optimisation. AI-enabled analysis of historical print data and consumption patterns can reduce energy waste by predicting future requirements, including device utilisation, automatically powering down during periods of inactivity, and load balancing. AI-enabled printer management software can drive efficiencies and generate energy savings by streamlining print workflows, job scheduling, routing, and making real-time print-setting decisions based on the printed document type to reduce waste.
• Supply chain optimisation. AI can improve the efficiency of print industry supply chains when used for demand forecasting, route optimisation, and inventory levels. These contribute to reducing energy, waste, and distribution and transport impacts. The vast amounts of data flowing across supply chains play to AI’s data-handling strengths.
• Waste production and management. AI can be implemented to drive improvements to product longevity, where products are specifically designed for the four Rs: repair, refurbishment, reuse, and recycling.
• Print consumables. AI can support the selection of more sustainable toners and inks by analysing their environmental impact and recommending greener choices, such as biodegradable inks. Added value can be gained by analysing usage data to reduce ink and toner consumption and using algorithms to adapt based on real-time data.
• Carbon offsetting programmes. AI can assist in calculating, monitoring, and validating carbon offset programmes, including tracking and verifying emissions reductions. This is important given criticisms around the integrity of offsetting and the limited near-term environmental benefits of some offset projects.
• End-user aids. AI co-pilots that adjust printer settings in real time can improve the end-user experience and quality of outputs. They can also minimise printing and paper usage and waste and support the move to hybrid paper/digital environments. The use of AI for cyber resilience should not be overlooked, as it can play a role in mitigating attacks, preventing unauthorised printing of sensitive and restricted materials, and protecting data.

Challenges and considerations

AI is still highly dynamic, and problems are being seen with its regular use. As such, it must initially be used to help with other approaches rather than attempting to totally replace them. As AI matures, safeguards can be turned off, allowing AI to make more decisions without the need for human checking.

Quocirca recommends that print manufacturers:

• Embrace open data standards. The foundation of effective AI implementation is a robust data infrastructure to manage, store, and process data to improve efficiency and decision-making. AI requires as much data as possible to conduct considered analysis. As such, the use of open data standards will allow the print environment to interoperate with the broader IT and business environment. AI models and training data are also critical. Cross-industry collaboration to create models that can be used across mixed printer fleets will help suppliers meet the carbon reduction ambitions of organisations.
• Availability of environmental data. Quocirca’s sustainability research reveals a major problem organisations face: limited data on the environmental impact of their print estates in terms of energy, emissions, and overall product lifecycle. This is another area that would benefit from collaboration across the print industry.
• Regulations and standards. These are critical for the safe use and reliability of AI-powered systems, but the global environment is a patchwork of evolving AI regulations, policies, and guidance.
• Ethical considerations. AI concerns about ethical use, safety and transparency, data privacy, and potential biases in algorithms are intensifying. Using AI to support sustainability creates another challenge. AI is under scrutiny because of the environmental impact of training and running models and queries within energy-hungry data centres, resulting in increased carbon emissions.
• Partnerships. Managing the complex landscape of data sources at scale and speed requires partnerships with hyperscale cloud computing providers such as AWS, Google Cloud, and Microsoft. They are AI technology leaders, so print suppliers should consider using their cutting-edge AI tools as the foundation for print-specific AI developments. Partnering with AI experts will accelerate AI initiatives and provide a way to utilise their expertise to identify AI-enabled opportunities.

Conclusion

AI is disruptive but can potentially transform the print industry over the next five years. The print industry can move towards a more sustainable future by harnessing its power and addressing its challenges. As sustainability is good for business – organisations are increasingly selecting suppliers on the basis of their sustainability credentials – suppliers can also boost their own performance. The scope to apply AI to sustainability initiatives is immense, with every aspect of the value chain, from design and engineering to aftermarket services, able to benefit.

It is important to be aware of the detrimental environmental impact AI can have due to its huge computing and energy demands and mitigate against it when deploying AI for sustainability. However, the fast pace of development means AI carbon footprint measurement is challenging. Print suppliers need to act quickly to grasp the opportunities AI can offer and partner to tap into technologies that will advance product and service innovation.