SAS: Optimising energy use in manufacturing with data and AI

SAS: Optimising energy use in manufacturing with data and AI


Now more than ever, manufacturers are looking for new ways to save energy. Optimising energy use was always important, now it is vital for the sustainability of the organisation.

The pressure is on to do more with less, while also maintaining quality and yield.

Cost is just one significant factor. Energy prices surged 60% in 2022, sparked by supply chain disruption and geopolitical tensions, and while those highs have fallen slightly, prices remain higher and are only likely to increase.

The International Energy Agency (IEA) has stated that the current global energy crisis is of ‘unprecedented depth and complexity’. 

Yet market rates for energy are only one ‘cost’ that CEOs, CFOs and COOs of manufacturing companies have to consider. Increased regulation around emissions, ESG requirements, public pressure, investor expectations – not to mention favourable financing available for ESG-compliant companies – all mean businesses have to do more to reduce the amount of energy that they use, and also switch to cleaner, more sustainable, energy sources.

This is a significant challenge, but also an opportunity.

In this White Paper, experts from SAS – a global leader in artificial intelligence (AI) and advanced analytics – share their insights and learnings from helping both process and discrete manufacturers globally to optimise their energy usage by implementing AI in a real world setting.

Key Energy Challenges in Manufacturing

The War in Ukraine is only one factor that has impacted energy prices. As the world emerged from COVID-19 restrictions in 2021, supply of energy was slow to meet surging demand. Gas, electricity and oil all hit record highs. 

When Europe’s over-reliance on Russian gas was exposed, and European Union nations tapped into LNG from other markets, that had a knock-on effect that drove up global prices as demand simply outstripped supply. It is basic economics in action.

Countries that had been turning away from fossil fuels and shutting down nuclear power plants were having to rethink their strategies due to energy security concerns – and these continue to impact energy prices.

The World Economic Forum’s Fostering Effective Energy Transition 2023 report states that after a decade of progress towards, clean, sustainable energy – those efforts have stalled as countries are shifting their focus towards energy security.

Businesses face the same challenges – securing energy at a manageable price point and maximising their efficiency to reduce total cost.

Cheap and stable energy tariffs are a thing of the past, which means that real cost optimisation requires dedicated action and more efficient and sustainable processes – especially for heavy industries with a high energy burden.

Part of that action is securing supply, especially for those heavy industries that still rely on non-renewable energy sources. The transition to clean, renewable energy is a long-term challenge – realistically 10 to 20 years – especially for these heavy industry manufacturers, so reducing consumption overall can assist in that transition.

Changing regulations around emissions are also driving the need to optimise energy usage and reduce an organisation’s carbon footprint. Failure to meet targets can mean heavy fines or even plant shutdowns.

In Europe, crucial pieces of legislation are in the pipeline, including the Green Deal Industrial Plan and the EU Net Zero Industry Act. These are driving the need for traditional industries to adapt and drive rapid change in how they consume energy and resources.

In the US, the Inflation Reduction Act of 2022 offers advanced manufacturing production credit and energy-efficient commercial building deduction to promote sustainability in business.

Additionally, company boards are making ESG commitments to shareholders, with bold statements on how they will become more sustainable. However, these often remain goals and rarely become actions – in part because it is difficult to break down such principles and goals to individual plants, production lines and processes.

Perhaps the biggest challenge manufacturers face is an internal one. With the ‘low hanging fruit’ of energy monitoring already picked, many plant managers and engineers may feel that there is no more that can be done when it comes to making their operations more efficient. Instead, they may look solely for cheaper energy procurement to deliver incremental gains.

When beginning an energy cost optimisation exercise across manufacturing operations, one of the main challenges that arise comes from trusting that advanced analytics and AI can do a good job. As engineers are typically conservative, working on operational excellence causes them to resist the analysis of a computer programme.

Manufacturers need to explore new ideas. Adopting advanced data analytics and using the latest AI software can help manufacturers optimise their energy usage, often delivering double digit cost savings by shining a light on ‘hidden’ opportunities.

The Analytics Opportunities

Many manufacturers are sitting on a data gold mine – the operative word being sitting. Companies need to take that data out of silos and bring it together in one unified analytics solution. Today, engineers are missing hidden insights that can improve upon process parameters by analysing the relationship between key metrics and the energy consumed. These insights will enable them to reduce energy consumption while maintaining or improving quality, yield, and throughput. 

Energy monitoring and optimisation can complement each other, particularly if data is yet to determine the greatest energy consumption and the potential areas for reduction. Manufacturers could see variations in energy consumption due to inefficiency of their production schedules or from fluctuating raw material specifications, changing ambient conditions, etc.

Not all manufacturers are created equal. Process industries, or energy-heavy industries are more aware of where the biggest energy consumers are. 

In discrete manufacturing, it must first be understood where the high energy consumption is, as well as the biggest variation in energy consumption – an area where there is great room for optimisation. 

Monitoring will provide key insights into energy consumption across a variety of operational factors, from equipment to administrative aspects, to determine an overall baseline. In the process, datasets will be formed to pinpoint energy-intensive processes. 

Advanced analytics helps customers analyse factors they may have overlooked. By deploying a prepackaged software solution in an agile way, manufacturers can see quick results. 

The Energy Optimisation Journey

For any transformation, it is essential to get buy-in from the C-suite to the factory floor. That is a challenge in itself.

A lot depends on the maturity of the manufacturer and willingness to trust in the data and AI solution. While the concept of AI has been with us for some time, it has recently entered the public consciousness as a perceived threat to jobs and livelihoods. This needs to be tackled, and AI embraced as a business tool that improves efficiency, profitability and, therefore, economic sustainability and viability.

Allowing engineers to experiment with the technology will help to build trust and deliver long-term benefits. Engineers may be sceptical of tweaks and solutions being suggested by the algorithms at first, but the data seldom lies.

Building this trust is essential to a successful and efficient adoption.

By combining a deep understanding of multiple variables, AI can simply find solutions that humans may fail to identify and improve through no fault of their own.

Manufacturing is complex, especially when it comes to variability in combustion processes that are energy-heavy. Being able to identify and mitigate variance between shifts or batches will deliver significant savings.

Not only that, being able to accurately measure and track progress is essential for sustainability benchmarking and reporting.

Advanced analytics avoids the disruption manufacturers would face if they were upgrading machinery or equipment to achieve efficiency gains. Using advanced analytics is faster and more risk-free than traditional transformation projects.

How SAS and AI Can Help

SAS recognises that AI has the ability to find the needle in the haystack and help leaders learn from past processes in a more scientific way. When compared to more traditional approaches like upgrading equipment to save energy, AI enables control of a number of variables to ensure the optimal settings for the machines at scale, very quickly, and with lower risk.

Furthermore, proof of concepts (POCs) are reliable for testing the effectiveness of energy cost optimisation products, but there are barriers to success. 

SAS has proven that the approach works if customers are ready to go along on the journey. SAS has the tools that enable customers to quickly connect the real data sources in a few weeks via cloud technology.

The best approach to minimise time to value and maximise the benefit is to look at how much energy can be saved. This should be the main driver. 

The role of data availability and data quality in a successful energy cost optimisation programme is to provide a backbone – this is the most important part. Unfortunately, for some customers who have a low degree of connectivity in their operations yet, it's also the most expensive part. But with use cases like energy cost optimisation, investment in connectivity really pays off.

It is important that insights are relevant and actionable for a process or industrial engineer to ingest. There is a strong need for process knowledge to ensure engineers are able to take data-driven actions on the factory floor. This will enable them to understand the process variables and the potential areas available for improvement.

When process engineers have a mindset of continuous improvement, one of the great skills they can learn and apply is root cause analysis to find underlying issues within processes and then implement appropriate corrective actions.

Transitioning to a data-driven culture, especially from a conservative starting point, requires acknowledging it as a journey rather than a sprint. This involves setting clear business objectives, gaining buy-in for the change process and sometimes modifying standard operating procedures.

However, it is important to start simple  – intelligent people often want to tackle the most challenging problems, but that can create a scenario where manufacturers try intellectually stimulating projects that have a long time to value while there are low hanging fruits that are much easier to harvest.

SAS says an energy cost optimisation pilot should be achievable in three months for an industrial pilot that already works 24/7. The main factor to consider is always data – if the data is not correct and prepared in the right way, it will take longer. 

A major part of SAS’s go-to-market strategy is to establish purpose-built ecosystems, which are designed to fulfil a customer’s need or solve a complex business problem in a very specific vertical. 

Microsoft, for example, has global energy optimisation as a strategic focus area. SAS has partnered closely with them, specifically on this topic.

SAS developed much of its energy cost optimisation technologies to be natively integrated with Microsoft’s Azure platform. SAS is also available on the Azure marketplace for manufacturers to purchase and deploy directly.

Wienerberger’s Energy Optimisation Success Story

Wienerberger AG is the world’s largest producer of bricks, with 217 production sites in 27 countries. The company is also the market leader in clay roof tiles in Europe as well as concrete pavers in Central-Eastern Europe and pipe systems in Europe.

Wienerberger was looking to optmise its energy usage, particularly in the brick kilns which require significant amounts of heat. They took the guesswork out of making improvements by combining and analysing IoT and process data from across the production process, using prebuilt analytic models to predict specific energy consumption and make sense of variable impacts – from quality of the clay to outside humidity and indoor temperatures. 

By using SAS software, Wienerberger delivered mathematically optimised insights that empowered production operators to tune the process recipe setpoints, accelerating time to value and providing an average energy cost saving of 10% – hugely significant savings for that manufacturing process.

WATCH the Wienerberger video


Find out how SAS can help you reduce your energy costs at

SAS: Optimising energy use in manufacturing with data and AI
SAS: Optimising energy use in manufacturing with data and AI
SAS: Optimising energy use in manufacturing with data and AI
SAS: Optimising energy use in manufacturing with data and AI