Brains of steel

Feb 19 2024

With modern sensor technologies, cloud connectivity and advanced analytics firmly in place, GEA is now adding a layer of intelligence to its hard-working machines. The results demonstrate the power of AI to take production processes to new levels of performance and sustainability.

While digitalization’s rapid advance has indeed pushed some “old world” hardware to the brink of irrelevance (think landline phones and DVD players), the machines that form the backbone of industrial plants and processes around the world are busier and more critical than ever. From food production to wastewater treatment, these steel separators, homogenizers, spray dryers, slicers, ovens, chillers, packaging machines and heat pumps – among many more – are asked to produce ever more output, while consuming less energy and generating less waste. To meet today’s demands, plant operators need their machines operating at unprecedented levels of productivity and efficiency. It’s a challenge, for sure, but they can take heart in one simple fact: the machines themselves are getting smarter.

The evolving intelligence of GEA machines goes back decades. In the early days, it was rudimentary sensor technology delivering isolated data around temperature, pressure or flow rates. Today, sensors are vastly more sophisticated and ubiquitous. The expansion of the Industrial Internet of Things (IIoT), together with more data-intensive machines and processes, means that machine builders like GEA now have tremendous potential to harness AI for developing new digital solutions for their machines.

GEA uses AI to add a layer of intelligence to existing sensor technology, data and analytics. An emerging generation of GEA machines is already out there making real-time adjustments based on detailed process design knowledge and understanding each customer’s operation. This opens the door for automated systems that can self-adjust while enabling continuous and maximum productivity. In short: AI is empowering GEA customers to get the most out of their machines and meet today’s unprecedented challenges, whether it’s talent scarcity, a hyper-competitive market or stricter climate and environmental standards. Simultaneously, GEA is aggressively exploring generative AI solutions to increase productivity across its own workforce to boost efficiency, promote innovation and support data-driven decision making across a range of disciplines from R&D, to sales, to service.

GEA machines 'smarten up'

Boosting the efficiency and environmental performance of GEA machines is the centerpiece of GEA’s climate strategy – and the company is wasting no time leveraging AI to make these machines smarter and more sustainable. The GEA OptiPartner solutions are designed to maximize productivity, offering customers a “full service” process optimization and maintenance package. Smart software combined with automation work together to adjust process parameters in real time while reducing the need for manual intervention. Working like an autopilot, GEA OptiPartner reduces variations, so machines operate closer to their maximum output, while giving operators complete visibility throughout the process and more time to focus on other tasks.

Applied to spray drying – a process used, among others, to create milk powder from liquid milk – GEA OptiPartner is achieving energy efficiency gains averaging 8 percent. Given the energy-intensive nature of spray drying, this translates into significant energy savings – equivalent to the energy utilized by 200 German households annually. GEA OptiPartner solutions are also ISO certified and carry GEA’s green, Add Better label, which means it is validated by TÜV Rheinland.

“GEA OptiPartner exemplifies the evolution towards an IT-based, process-optimization-as-a-service, which leverages our extensive process design knowledge and understanding of each customer’s operational activities,” says Dr. Hassan Yazdi, Director, Program Management Digital Solutions, GEA. “GEA becomes a true partner in the production process: identifying areas for improvement, reducing unscheduled shutdowns and operating as an inclusive part of the process optimization team.”

Currently deployed in wastewater treatment, GEA Intellicant is GEA’s breakthrough in the digitalization of decanters – an AI-powered, cloud-based digital solution already available to customers today. The system consists of a sensor package plus GEA software known as virtual machine operator (VMO). The sensor package monitors three critical process parameters in real time: the concentration of solids in the wastewater intake, the turbidity of the dewatering liquid (centrate) and the dry substance of the dewatered sludge (cake). From these “sensory organs” data is transmitted to the VMO software – the “brain” – which reacts automatically and intelligently to the incoming data.

Because the makeup of the incoming sewage sludge entering the decanter is always changing, the decanter needs continual adjusting to achieve optimal performance. “GEA Intellicant solves a major problem for wastewater plants by allowing for constant, optimal operation of decanters under varying conditions – something that would be nearly impossible if operated manually,” explains Christian Schramm, a mechanical engineer by training and today Product Manager Commercial for Digital Products at GEA.

GEA Intellicant is already helping wastewater plants in Germany and Denmark achieve significant efficiency gains in sludge dewatering to minimize waste and improve cost efficiency. “With GEA Intellicant, a municipal water treatment plant (serving 125,000 population equivalent) can reduce the amount of sewage sludge by some 400 metric tons per year. That’s not only good for the environment, but also reduces disposal costs by up to 60,000 euros per year,” says Schramm.

The GEA InsightPartner Brewery solution is another example of AI enabling real-time responsiveness and flexibility. In brewing, GEA InsightPartner integrates predictive AI elements into its real-time monitoring solution, providing operators with optimization suggestions.

Instead of waiting for one brew to finish before adjusting for the next, the system also allows operators to improve processes mid-brew – for instance, by fine-tuning energy recovery as the wort boils, or by correcting imbalances in the energy storage system. “Here, too, the responsiveness afforded by AI paves the way for a significant change in approach: away from ‘recipe based’ in favor of flexible condition brewing,” says Dr. Mark Schneeberger, Senior Director Product Development, Design & Research, GEA.

GEA InsightPartner Brewery allows brewers to understand causality and correlations between data thanks to advanced analytics and a well-presented dashboard. In addition to conventional brewhouse data, the software also analyzes sustainability performance indicators based on energy consumption and heat recovery in the process cycle. So far, users have reported productivity gains of between five and eight percent.

Working with partner CattleEye, an agribusiness based in Northern Ireland, GEA offers an AI-powered solution to quickly detect and treat lameness in cows. CattleEye consists of a camera installed above the guided exit of a milking system that captures and records each cow as it leaves the milking parlor. The associated software analyzes animal movement patterns, evaluates irregularities, applying a mobility score to each animal in the system. Via the app, farmers can quickly identify sick or lame cows which may need treatment.

“Claw diseases and associated lameness are the third most common cause of dairy cow deaths – and thus a major risk factor for farms overall,” says Holger Siegwarth, Vice President Digital Solutions and Services at GEA. “With CattleEye, GEA customers can leverage AI to keep a closer eye on the hoof health of the herd and take appropriate countermeasures. This not only makes farms more profitable, but also allows them to operate more sustainably. Animals are healthier overall and live longer."

Laying the groundwork

With individual AI solutions already delivering real value to customers in the critical fields of wastewater treatment, food, beverage and dairy, GEA is moving quickly to integrate AI more broadly in its machines. In January 2024 it launched GEA X Control, a new control system for centrifuges that will make data collection and analysis faster and easier, paving the way for customized software functions as well as self-learning and -optimization capability. GEA X Control takes advantage of the latest microcontroller and processor technology to achieve new levels of connectivity, allowing users to adapt their processes more quickly and efficiently to changing market needs. This solution is available for selected separator and decanters used in the dairy and renewables industry.

Generative AI at GEA: turning hype into action

The world is still buzzing about ChatGPT and how generative AI will forever change the way we learn, work and create. A recent McKinsey study suggests that traditional AI – such as the edge/cloud AI services propelling GEA machines into the future – will have the largest economic impact moving forward. While generative AI is expected to deliver significant “on top” benefits for companies internally – well beyond just text and image generation.

“Generative AI plays a vital role in securing GEA’s competitiveness,” explains Tom Oelsner, Chief Digital Officer at GEA. “GEA will seize the opportunities of generative AI in areas such as AI-powered software development, generative machine design, intelligent technical documentation as well as in redefining sales and service processes.”

To help realize the potential of generative AI, GEA has developed its own company-wide chatbot GEA BetterBot AI, which is based on OpenAI’s ChatGPT as well as Microsoft’s Azure OpenAI Service. Launched in 2023, GEA BetterBot AI is fed with GEA-internal data and runs securely in the GEA Cloud. This means employees can draw on GEA-related information within seconds without worrying about confidential data getting out. “The chatbot can help staff respond quicker to service requests from customers, access market intelligence data or support marketing activities,” says Oelsner.

To help employees build their chatbot “communication skills”, the GEA Academy developed a tutorial on prompt engineering. The Academy offers a host of learning opportunities – everything from digital product development to e-commerce skills, data science techniques and agile project management.

The AI difference

As advances in both traditional and generative AI continue to grab the headlines – and CEOs everywhere are urged to keep up with developments – GEA is cutting through the hype, taking concrete steps to integrate machine intelligence into the everyday business of engineering. From optimizing product development, marketing and service processes, to equipping GEA machines with powerful new digital solutions, to testing new boundaries of sensor technology, edge computing and intelligent automation, GEA is leveraging AI to achieve levels of standardization, consistency, efficiency and sheer speed of execution that would be hard to reach without its help. As the climate crisis heats up and pressure mounts from stakeholders across the board, GEA demonstrates how AI can make a real difference for GEA and its customers – and help engineer for a better world.