AIRBUS AND Ericsson have successfully deployed a private 5G solution at the Airbus production site in Hamburg, with another deployment underway in Toulouse. This initiative forms part of Airbus’ ambitious digitalization strategy, aimed at strengthening manufacturing automation, traceability, and operational efficiency, while meeting the sector’s strictest safety and security standards. 

The partnership between Ericsson and Airbus leverages Ericsson Private 5G, recognized for its reliability, security, and high performance. The solution’s built-in infrastructure automation enabled rapid deployment across Airbus’ operations, significantly shortening implementation timelines compared to traditional setups. This automation allowed Airbus to scale connectivity quickly and securely across multiple sites. Close collaboration with the Ericsson product team ensured seamless integration, with the solution tailored to Airbus’ IT-tooling and cybersecurity requirements. The design’s modular architecture and API-driven interfaces simplified onboarding into Airbus’ existing systems, accelerating time-to-value and reinforcing robust security controls.  

Private 5G network

With a fully operational private 5G network now live in Hamburg and deployment at Toulouse underway, this rollout is part of a broader roadmap to extend private 5G across Airbus’ strategic sites in Europe, including further locations in Spain, the United Kingdom, and internationally, with projects in the United States pending.  

This effort reflects Airbus’ commitment to standardizing digital operations and scaling innovation across its global footprint. 

Hakim Achouri, 5G expert at Airbus, says: “Our objective is to migrate all our industrial networks towards 5G to ensure unified, ultra-reliable connectivity from the operator’s workstation to the aircraft cabin. This deployment accelerates projects involving 3D simulation, augmented reality, improved traceability for parts, and predictive maintenance for our assets. The standardization and scalability made possible by this architecture allow us to replicate the solution easily across further sites in Europe and worldwide.” 

Manish Tiwari, head of enterprise 5G, enterprise wireless solutions, Ericsson, says: “Our collaboration with Airbus embodies the alliance between technological innovation and  industrial excellence. Ericsson is proud to support Airbus’ digitalization ambitions through Ericsson Private 5G, offering best-in-class, secure connectivity at scale.” 

IoT integration

Ericsson Private 5G forms the backbone of Airbus’ strategic transformation projects, enabling high-value industrial use cases such as Internet of Things (IoT) integration, intelligent management of critical equipment, real-time quality control, and collaborative robotics. With seamless, full-site coverage with private 5G, machines and operators on the production floor gain true mobility, boosting productivity, process agility, and end-to-end industrial control, all of which are key to realizing the full potential of Industry 4.0.  

This new phase underscores Airbus and Ericsson's commitment to the future of industrial connectivity, featuring advanced 5G Standalone (SA) technology and next-generation deployment models, which are also poised to accelerate 5G usage in office environments. Additionally, joint R&D efforts focus on connected cabins, 6G, and nonterrestrial networks (NTN), enhancing the connectivity ecosystem for aerospace and smart manufacturing applications. Through this strategic partnership, Airbus and Ericsson are accelerating the digital transformation of the aerospace industry, laying the foundation for the next generation of smart factories — fully connected, scalable, and sharply focused on innovation across Europe and the world. 

Find out more about similar projects by visiting Stand K124 and the Connected Production Theatre, where Ericsson will be presenting rhought the two days of the exhibition.

www.ericsson.com

VISIBILITY GAPS remain a major obstacle in industrial cybersecurity. In its 2026 OT threat reporting, Dragos estimates that "fewer than 10% of OT networks worldwide currently have meaningful network monitoring in place", leaving defenders with limited insight into malicious activity until an incident is already underway.

But simply bolting traditional IT visibility tools designed for servers and endpoints into OT networks can, and often does, create instability or degraded performance. To protect cyber and operational resilience, manufacturers need an engineering-led view of network insight that respects uptime and live production.

The visibility paradox

Traditional IT security tools often rely on active scanning or inline inspection, methods that can create latency in fragile control systems if they are used carelessly.

Take industrial controllers supporting live production processes. These systems rely on real-time communications to keep machinery operating within expected tolerances. Unexpected scans or intrusive network testing can introduce delays or disrupt those communications, which in turn can affect output, quality or availability.

Paradoxically, organisations cannot secure what they cannot see. Yet attempting to observe these networks using conventional IT methods can destabilise the very systems they are trying to protect.

As NIST states in its Guide to Operational Technology (OT) Security, "OT network owners should exercise extreme caution when permitting active scanning on an operational network due to device sensitivity on the target network. Active scans may cause device instability or interfere with the device process state, potentially impacting safety and integrity."

Passive monitoring resolves this. By observing network traffic through engineered SPAN or TAP connections, it gives operators a way to understand communications without interacting directly with sensitive devices. That makes it better suited to fragile environments where active scanning may introduce operational risk.

In live manufacturing environments, these approaches should be designed and validated before they are rolled out. Passive monitoring across plant networks can support continuous asset discovery and exposure analysis without adding traffic to production systems. That helps replace incomplete manual inventories with a clearer picture of operational assets and communications.

Scalable resilience

Visibility alone does not reduce risk unless it informs how networks are structured and governed. In many organisations, OT systems that run machinery are connected to the wider business network without enough planning or separation.

That creates unnecessary exposure. Once a business network is compromised, attackers can move more easily towards critical operational systems. Legacy OT devices often cannot support modern security agents or deep packet inspection, which leaves them particularly exposed when networks are merged without clear boundaries.

The next step is turning visibility into controlled, resilient infrastructure. As the NCSC notes in its secure connectivity principles for OT, connectivity "should be designed with operational resilience in mind, and should not compromise the safety, reliability, or availability of OT systems."

The progression is straightforward. Organisations first need to identify connected assets and communication flows so they understand how systems behave under normal conditions. Segmentation can then be introduced through methods such as VLANs and network isolation to separate domains according to operational importance or trust level. Continuous monitoring then helps ensure those boundaries remain effective over time.

In manufacturing environments, network security should be embedded into the architecture from the outset. Manufacturers looking to strengthen resilience across industrial environments should start by asking not just whether they can see their environments, but whether they can do so safely and continuously.

Carl Henriksen is CEO at OryxAlign

www.oryxalign.com

The company, which is headquartered in Leek, the Netherlands and expanded into the UK in 2025, is also celebrating its 20th anniversary as a leading provider of refurbishment and repair services for industrial electronics. 

The Buyback scheme enables businesses to dispose of surplus, obsolete or faulty industrial electronics free of charge, while also generating revenue from equipment that might otherwise be discarded. 

Obsolescence in industrial products has long presented a challenge, often leading to unnecessary waste when older components are discarded prematurely. Instead of Waste Electrical and Electronics Equipment (WEEE) compliant disposal fees, businesses can now generate revenue from equipment that would otherwise be thrown away. 

The JC-Electronics’ experienced team will visit customer sites to assess surplus stock, provide guidance on long-term equipment support and offer a competitive price for viable products. This removes the need for organising skips or arranging disposal, helping companies to reduce costs and streamline operations. 

Visitors to Smart Manufacturing Week can enter the ‘Golden Ticket’ initiative at the JC-Electronics stand G110 for a chance to win £500 off their next repair. No purchase is necessary and the prize can be applied to any repair service. 

Dan Jones, buyer at JC-Electronics, said: “This new scheme encourages companies to explore alternative ways of managing obsolete machinery, supporting more sustainable supply chains while ensuring compliance with WEEE regulations.”

With a team of 140 specialised technicians, JC-Electronics repairs, refurbishes, calibrates and tests every product to the highest standard to achieve the JCertified warranty seal. This supports effective management of the industrial electronics supply chain while delivering sustainable and cost-effective automation solutions. 

Jones continues: “We want to demonstrate that there is a viable alternative through the use of refurbished industrial automation spare parts. For the past 20 years, by choosing JCertified refurbished electronics, customers have reduced their consumption of water, raw materials, CO2 emissions and benefited from a two-year warranty.” 

jc-electronics.com/uk

CRITICAL FOR passenger safety and comfort, lift and elevator controllers also increasingly rely on digital and connected technology for energy efficiency and remote condition monitoring, driving the need for advanced control solutions. 

Peter Roberts, project manager at manufacturer of traction and hydraulic lift and escalator controllers and accessories, Lester Controls, says: “Our technology provides accurate positioning and speed of lifts and elevators, and meets all the strict safety regulations in the event of a power failure."

"We’re also responding to the industry’s trend towards digital and energy-efficiency, which doesn’t have to be tied only to high-cost investment of new infrastructure. For lift and escalator applications, both new and existing, controllers must be digitally connected, sending data to the cloud via the internet or mobile phone network. Maintenance engineers use this information to diagnose faults remotely and arrive on site with the right spares and tooling. That cuts operational costs by reducing the number of site visits."

"Another popular feature is energy-saving. We can automate escalators to go-slow or stop at quiet times, and we often add energy recuperation or variable speed control for lifts."

A particular challenge for Lester Controls is that every building is different, with lift control logic varying depending on the number and type of lifts and escalators. Each project needs a dedicated controller to be designed, built, and tested. This is a complex process that requires accuracy over every detail. That is why Lester Controls has adopted ETAP SEE Electrical as the CAD package it uses to create electrical schematics.

The challenge

When designing new lift and elevator controllers, there are many variables at play. These include the number and types of lifts and elevators, the control philosophy, energy-saving features, and requirements for remote monitoring and communications. 

All of these requirements lead to unique functionality for each project. In turn, this requires a unique arrangement of components such as PLCs, motors drives, switches and HMIs for each project. Lester Controls’ engineering team uses specialist CAD software to design and manufacture its controllers.

A cabinet might contain up to 160 individual components. Any error can lead to delays in the workshop, as production engineers need to raise queries and resolve issues when they are building the cabinets. These errors can be avoided with modern software, which includes features such as contact registers. 

As Roberts explains: "A customer got in touch in 2020 to request that we use specific symbols in our drawings. As these were standard symbols that are in the latest technical standards, we realised it was time to upgrade to a modern intelligent CAD package."

The solution 

"We evaluated all of the software packages on the market and came to the conclusion that ETAP SEE Electrical was best for our needs. It’s a software package from Schneider Electric, a collaborative supplier of components for our controllers and a long-standing member of our ecosystem. It comes with technical support and training resources, has flexible licensing and supports all the symbols in the latest standard."

"We wanted an intelligent package that would enable us to reuse the data from drawings. For example, we need to export parts lists in CSV format for upload into our parts management system, and we needed it to integrate with our existing printer for identification labels. We also needed to interface with our cutting machine, as well as our manufacturing resource planning (MRP) software."

"Another important feature was an automated register of electrical contacts. With old CAD package, we had to reconcile these individually, which was painstaking work. Any loss of focus could result in duplicate contacts in a drawing, leading to technical queries from the workshop, taking more time to resolve."

"After installing the software, I picked up the task of reworking all of our existing drawings in the new software. Balancing it alongside project work, it only took eight months to convert around 200 master drawings and 800 symbols. I found the new software easy to pick up with tutorial videos, PDF guides, and support from Schneider Electric."

"Since then, it’s been fairly easy to roll out to the team, which includes seven designers and three technical staff. We also brought on a new designer, who learned it after seeing the basics and getting a couple of drawings. Nobody in the team has had major questions, it has worked flawlessly and has future-proofed operations as new staff are brought onto the workforce."

"The big result is that the software has saved nearly half of our design time. Previously, a major contract would require about five days of design and a lot of manual data entry but now it’s three days. Typically, a major project has around 160 components and we’d need to manually paste part numbers and descriptions to create a parts list that corresponds to each drawing. But now we can simply extract the data from the software as a CSV file."

"The software has also reduced the number of design errors. Because it includes a register of contacts, we no longer get duplicate contacts. That means we have fewer technical queries to resolve with the workshop, saving more time and offering more value to customers."

etap.com/home

According to the court’s decision, Elite Robots Germany is immediately prohibited from offering or distributing the infringing software and all products containing this software in Germany until further notice.  Moreover, Elite Robots Germany is obligated to provide comprehensive information about the infringing acts it has committed and, in doing so, also to disclose information about the customers it has supplied. Teradyne Robotics intends to take legal action against Elite Robots’ distributors and partners if they continue to offer the infringing software.

“At Teradyne Robotics, we have chosen to take a stand against any competitors copying our proprietary hardware or software design and we are of course pleased with this ruling,” said Jean-Pierre Hathout, president of the Teradyne Robotics Group. “We believe we have irrefutable evidence of copyright infringement and, while this is not a final ruling from the court, it is a clear indication that we have a very strong case.”

Hathout adds: “Automation and innovation are critical to our industrial future. We cannot passively allow companies to unlawfully copy protected technologies. This not only hampers research and innovation but also undermines customer experience and confidence. Teradyne Robotics remains fully committed to protecting our intellectual property and to ensuring automation customers have access to the safe, innovative and high-quality solutions they deserve.”

www.teradyne.com

INDUSTRIES TODAY face a complex mix of pressures that make sustainable automation essential. Environmental responsibility is among the most visible. Industrial operations contribute significantly to global energy consumption and carbon emissions, and automation systems, particularly in manufacturing, assembly and logistics, are often energy-intensive. Material waste is another pressing concern. Scrap, defective products, and packaging inefficiencies increase costs and place additional strain on the environment. Sustainable automation addresses these challenges while delivering measurable economic benefits.

Rising energy costs and market volatility add urgency. Organisations that invest in energy-efficient equipment and process optimisation can achieve significant cost reductions. For example, a mid-sized automotive facility that replaced standard motors with high-efficiency variable-speed alternatives reduced electricity consumption by 20%, resulting in annual savings of more than £400,000. Process optimisation aimed at reducing material waste can have a substantial impact, particularly in industries where raw materials account for a large share of production costs.

Regulatory requirements further accelerate adoption. Across the UK and EU, carbon pricing mechanisms, mandatory energy reporting and emissions reduction targets establish clear expectations. Compliance with standards such as ISO 50001 provides a structured approach to improving energy performance, while transparent reporting strengthens stakeholder confidence. Organisations that address sustainability proactively can reduce regulatory risk, strengthen credibility and in some cases access financial incentives for early action.

Principles of sustainable automation

Sustainable automation rests on four key principles: energy efficiency, waste minimisation, lifecycle sustainability, and intelligent process management.

- Energy efficiency is foundational. Automated systems should operate only when required, and equipment such as motors, drives, conveyors and robotics must be correctly sized and configured to minimise power consumption. AI-driven scheduling and regenerative technologies can further reduce energy demand. In one industrial automation project, the introduction of regenerative robotic arms, combined with AI-based scheduling, reduced energy use by more than 20%, delivering annual savings of £350,000 and significantly lowering carbon emissions. When applied at scale, decisions like these have a substantial cumulative effect.

- Waste minimisation focuses on preventing scrap, rework, and inefficient use of materials. Robotics and precision automation improve accuracy, while sensors and analytics detect inefficiencies before they escalate. Closed-loop recycling systems, where scrap materials are reintegrated into production, further reduce waste. In electronics manufacturing, precision robotic assembly reduced defective product rates by 30%, cutting material waste and lowering disposal costs.

- Lifecycle sustainability requires organisations to consider environmental impact from system design through to end of life. Selecting recyclable materials, designing equipment for maintainability and planning responsible disposal ensures sustainability is addressed from the outset rather than added later. Predictive maintenance extends equipment lifespan and reduces unnecessary replacement, delivering both environmental and financial benefits.

- Intelligent process management ensures continuous improvement. IoT sensors and digital twins enable real-time monitoring of energy use, material consumption, and machine performance. AI-driven analytics identify inefficiencies and recommend operational adjustments before issues emerge. In food processing environments, AI-based production scheduling reduced energy consumption during peak periods while lowering spoilage, demonstrating how intelligent management can improve efficiency and sustainability simultaneously.

Enabling technologies

Numerous technologies enable sustainable automation, each contributing distinct capabilities. However, the greatest gains are achieved when these technologies operate as an integrated system rather than in isolation.

Robotics and smart manufacturing systems improve accuracy, reduce human error, and lower material waste. Energy-efficient drives, regenerative technologies, and intelligent scheduling reduce power consumption, while integrated control systems ensure different parts of the operation communicate effectively, limiting inefficiencies across production and logistics.

Artificial intelligence and machine learning play a critical role in predictive maintenance, process optimisation and demand forecasting. Predictive maintenance reduces unplanned downtime and extends equipment life, while process optimisation supports more energy-efficient production. Accurate demand forecasting aligns output with actual market requirements, helping to reduce excess inventory and material waste. In a large packaging facility, AI-based scheduling reduced peak-hour energy consumption by 18% and lowered spoilage by 12%.

Industrial IoT provides the real-time data required for informed decision-making. Sensors monitor energy usage, machine performance and environmental conditions, generating actionable insights. By analysing this data, organisations can identify inefficiencies, prevent waste and continuously optimise production processes.

Renewable energy integration complements automation technologies by reducing reliance on grid electricity. Solar, wind and biomass systems can supplement or replace traditional energy sources, lowering carbon emissions. Smart energy management systems coordinate renewable energy supply with production schedules. This helps maintain operational continuity while maximising the use of low-carbon power.

The common thread across these technologies is integration. Robotics alone does not reduce energy consumption and renewable energy alone does not eliminate waste. Meaningful results come from connecting systems, processes and people so they operate as one.

Strategic recommendations

For leaders considering investment in sustainable automation, the path forward can be summarised into a few key actions.

- Start with insight, not technology. Organisations should begin by auditing energy use, waste and operational performance to build an accurate view of where improvement will deliver the greatest value. This clarity helps prevent misallocation of capital and ensures efforts are focused on areas with the strongest impact.

- Design holistically. Automation should be approached as a connected ecosystem rather than a collection of isolated assets. Software, hardware, energy, maintenance and people must work together. Integration should be a design principle, not an afterthought.

- Prioritise lifecycle value. Decision-making should extend beyond initial purchase cost to consider total cost of ownership. Maintenance requirements, energy consumption, operational lifespan, recyclability and carbon impact all influence long-term value. In many cases, the most sustainable option is also the most financially resilient.

- Invest in people. Sustainable automation depends on skills as much as technology. Developing capability in energy management, automation systems, data analytics and sustainability leadership strengthens execution. When teams understand the purpose behind the technology, adoption improves and outcomes become more consistent.

- Build strong partnerships. No organisation delivers sustainability in isolation. Collaboration with technology providers, integrators, energy specialists, academic institutions and supply chain partners enables access to broader expertise and shared learning, leading to more effective outcomes.

- Measure, review, improve. Sustainability performance is not static. Continuous measurement, review and adjustment are essential. Embedding sustainability metrics into day-to-day operations ensures progress is monitored consistently rather than treated as an annual reporting exercise.

- Embed cultural ownership. Sustainability should form part of organisational identity. Recognising success, sharing results and encouraging innovation at all levels fosters shared ownership. When sustainability is embraced collectively, it becomes a sustained capability rather than a top-down directive.

When organisations follow these principles, sustainability moves beyond cost containment. It becomes a strategic capability that supports long-term resilience, performance and value creation.

Dan Migliozzi is sales director for the UK, EU and North America at AGITO Global

agitoglobal.com

 

The new facility has been designed as a structured environment where the full portfolio of HaiPick Systems, including the company’s latest upgraded HaiPick Climb system, can be assessed under real operating conditions. By bringing together live HaiPick systems, engineering expertise, and project experience in one location, the centre enables more practical discussions on system design, scalability, and deployment across a range of industries including e-commerce, apparel, grocery, and industrial applications.

“The innovation centre provides a space for more informed conversations around automation,” comments David Burggraaf, marketing director EMEA at Hai Robotics. “Rather than focusing on specifications alone, it allows customers and partners to evaluate how systems perform in real operational scenarios and how they can be adapted to specific application needs.”

The facility has already hosted several technical sessions with industry partners prior to its official opening and will continue to serve as a hub for customer engagement, partner collaboration, and industry dialogue across the EMEA region. 

With increasing demand for flexible and scalable automation solutions in Europe, the innovation centre reflects Hai Robotics’ commitment to supporting customers throughout the design and deployment process, from early-stage evaluation to long-term system expansion.

www.hairobotics.com

Image: Agratas

Agratas' project will strengthen economy security and reduce Britain's reliance on imports by turbocharging domestic battery production and generating around £43 billion worth of economic growth over a 25-year period when the facility is in full operation.

The site will not only support 4200 direct jobs but thousands more in the supply chain, as well as unlock 300 apprenticeships – backed by a specialised battery manufacturing training unit to meet the skills needs of Agratas' gigafactory and the wider battery sector.

During a vist to the company, business secretary Peter Kyle (pictured) said: “This government is backing the industries of the future by investing in auto firms, SMEs and battery manufacturers across the country, helping to boost economic growth and our resilience, secure jobs and put more money in people's pockets.

“In an unstable world, our Modern Industrial Strategy is providing investors the stability and confidence they need to plan not just for the next year, but for the next 10 years and beyond. That is what sets us apart from the rest, and will help ensure advanced manufacturing remains a thriving sector in the UK for decades to come.” 

Earl Wiggins, vice president of manufacturing operations, UK for Agratas said: “We welcome the UK Government’s investment as we build a battery manufacturing facility that will play a vital role in delivering net zero and strengthening the UK’s position as a global leader in battery manufacturing.

“This funding will support the development of our Somerset facility, enabling us to produce battery cells for our anchor customer, JLR (Jaguar Land Rover). Over the next year we will have over 2200 people working on the site, and that growth will continue over the coming years." 

The latest Quarterly Update reveals that since the launch of the Modern Industrial Strategy over £360 billion of private investment has been secured across its key sectors, supporting up to 120,000 jobs. Alongside this, the government is cutting electricity costs for energy-intensive manufacturers, reducing unnecessary planning delays and overhauling regulation that holds back our most ambitious businesses.

Government is injecting £47 million worth of support for key R&D battery projects through the Battery Innovation Programme, helping to create skilled jobs, a stronger supply chain and position the UK as a globally competitive destination for battery manufacturing.

Auto businesses will also benefit from a £190 million boost to ensure the automotive industry remains ahead of the competition on the global stage. Startups and well-established firms including Nissan and Jaguar Land Rover have been awarded £90 million in DRIVE35 funding to ramp up innovative prototype and cutting-edge projects, strengthening firms’ technological capabilities and improving the affordability of EVs for customers.

www.gov.uk/government/organisations/department-for-business-and-trade

In a similar vein, Total Productive Maintenance (TPM) has followed a very similar trajectory. Originating in Japan in the 1970s, this philosophy encouraged all operators to take a proactive role in machine care and introduced the concept of preventive maintenance. Now, the modern descendant of TPM, Overall Equipment Effectiveness (OEE), is a key shop-floor metric worldwide, producing data that can feed smart automation even further. This symbiotic relationship enables operators to create a reactive, self-improving production line.

The what and why of OEE

Stripped to the studs, OEE is a holistic way of answering what is, fundamentally, a pretty basic question: Is this asset operating at its full potential? 

Uptime and output can help answer that question in part, but they aren’t all-encompassing indicators of function. OEE provides a more holistic solution, built on the three dimensions of availability, performance and quality. 

• Availability: assesses whether equipment is running when it is supposed to. Downtime and stoppage, both planned and unplanned, are factored into this metric. 
• Performance: dives into the oftentimes hidden inefficiencies of the machine. A breakdown is easy to notice, but a drop in speed due to a misconfiguration can slip under the radar and accumulate costs over time. 
• Quality: captures how much of the output meets the standard. Defects and reworks both impact the quality score of a given piece of equipment, whether due to wear or input errors. 

Taken together, these metrics paint a clearer picture of equipment health than any single figure, but an OEE score is still not fully indicative of efficiency. For all its strengths as a single, universal percentage, the final figure can still obscure the complexity of the losses it measures. For example, micro-stoppages tend to slip through the cracks without triggering alarms, despite their cumulative impact on performance. 

In many facilities, OEE is still collected manually through end-of-shift batch reports, a process that entails several limitations. Manual logging introduces a margin for error, neglects finer details such as timestamps, and, most importantly, only identifies issues after they have already occurred.  Enhancing the value of this final OEE figure and addressing its limitations involves moving from retrospective analysis, to a live operational system. 

Making automation truly smart

OEE, paired with real-time data, becomes a driver of decision-making. This focus on intelligent, interconnected networks that exchange information and automatically adjust operations is central to the modern idea of the smart factory, much like how transport software solutions optimise real-time decision-making across logistics networks.

Within this ecosphere, OEE acts as a feedback loop between machines, systems, and operators. When combined with tools such as Manufacturing Execution Systems (MES) and SCADA platforms, the Overall Equipment Effectiveness data enables:

• Automated alerts: Immediate notifications when any OEE metrics fall below approved thresholds, along with predefined escalation protocols to ensure the discrepancy is addressed as soon as possible. 
• Dynamic adjustments: Operators or automated systems can rebalance and reroute workloads, or adjust speed, to compensate for production drops in real time. 
• Root cause visibility: Data highlights exactly where and when issues occur, dramatically cutting investigation time.

Closed-loop controls, such as those enabled by real-time OEE data, have been shown to improve the efficiency and responsiveness of production systems. Cyber-physical manufacturing, as envisioned in “Industry 4.0”, creates an environment in which automation continuously monitors and improves the production process.  

Reducing downtime and driving capacity

Perhaps the most immediate benefit of real-time OEE is the ability to optimise productivity from top to bottom. Small inefficiencies, such as start-up delays and micro-stoppages that are typically unseen by traditional OEE, are detected immediately. Managers no longer have to dedicate time and resources to frequent manual inspection, another hidden waste that adds up. 

Data draws attention to these losses by identifying performance patterns that manufacturers can address without additional investment. This is precisely why modern strategies emphasise leveraging real-time insights to reduce downtime and maximise OEE: it’s a cost-effective, calculated approach to eliminate inefficiencies.

A second, equally impactful benefit of real-time OEE data lies slightly beyond the present. Predictive and automotive maintenance allow operators to anticipate failures based on performance trends, trigger maintenance, and adjust capacity elsewhere to compensate for the stalled output, with minimal manual intervention. 

These “self-optimising systems” are becoming more accessible and effective thanks to advancements in AI and greater integration of smart systems in factories. 

From insight to sustainable impact

Making automation truly smart through OEE data is not without its challenges. For all of these systems to communicate, they must first be compatible, meaning any setting still using legacy systems or non-digitised machines faces a hefty upfront investment. Thankfully, this process can be gradual, giving manufacturers a chance to identify the right integration strategy and to encourage operator involvement, in line with the TPM philosophy. 

These barriers are well worth navigating, as industry trends move towards a shared operational language built upon data visibility. By embedding real-time OEE into everyday workflows, manufacturers are reducing inefficiencies, speeding up maintenance, and unlocking the full potential of their existing assets.

Image: MIT

To avoid such an avalanche of inefficiencies, researchers from MIT and the tech firm Symbotic developed a new method that automatically keeps a fleet of robots moving smoothly. Their method learns which robots should go first at each moment, based on how congestion is forming, and adapts to prioritise robots that are about to get stuck. In this way, the system can reroute robots in advance to avoid bottlenecks.

The hybrid system utilises deep reinforcement learning, a powerful artificial intelligence method for solving complex problems, to figure out which robots should be prioritised. Then, a fast and reliable planning algorithm feeds instructions to the robots, enabling them to respond rapidly in constantly changing conditions.

In simulations inspired by actual e-commerce warehouse layouts, this new approach achieved about a 25 percent gain in throughput over other methods. Importantly, the system can quickly adapt to new environments with different quantities of robots or varied warehouse layouts.

“There are a lot of decision-making problems in manufacturing and logistics where companies rely on algorithms designed by human experts. But we have shown that, with the power of deep reinforcement learning, we can achieve super-human performance. This is a very promising approach, because in these giant warehouses even a 2 or 3 percent increase in throughput can have a huge impact,” says Han Zheng, a graduate student in the Laboratory for Information and Decision Systems (LIDS) at MIT and lead author of a paper on this new approach.

Zheng is joined on the paper by Yining Ma, a LIDS postdoc; Brandon Araki and Jingkai Chen of Symbotic; and senior author Cathy Wu, the Class of 1954 Career Development Associate Professor in Civil and Environmental Engineering (CEE) and the Institute for Data, Systems, and Society (IDSS) at MIT, and a member of LIDS. The research appears today in the Journal of Artificial Intelligence Research.

Rerouting robots

Coordinating hundreds of robots in an e-commerce warehouse simultaneously is no easy task.

The problem is especially complicated because the warehouse is a dynamic environment, and robots continually receive new tasks after reaching their goals. They need to be rapidly redirected as they leave and enter the warehouse floor.

Companies often leverage algorithms written by human experts to determine where and when robots should move to maximise the number of packages they can handle.

But if there is congestion or a collision, a firm may have no choice but to shut down the entire warehouse for hours to manually sort the problem out.

“In this setting, we don’t have an exact prediction of the future. We only know what the future might hold, in terms of the packages that come in or the distribution of future orders. The planning system needs to be adaptive to these changes as the warehouse operations go on,” Zheng says.

The MIT researchers achieved this adaptability using machine learning. They began by designing a neural network model to take observations of the warehouse environment and decide how to prioritise the robots. They train this model using deep reinforcement learning, a trial-and-error method in which the model learns to control robots in simulations that mimic actual warehouses. The model is rewarded for making decisions that increase overall throughput while avoiding conflicts.

Over time, the neural network learns to coordinate many robots efficiently.

“By interacting with simulations inspired by real warehouse layouts, our system receives feedback that we use to make its decision-making more intelligent. The trained neural network can then adapt to warehouses with different layouts,” Zheng explains.

It is designed to capture the long-term constraints and obstacles in each robot’s path, while also considering dynamic interactions between robots as they move through the warehouse.

By predicting current and future robot interactions, the model plans to avoid congestion before it happens.

After the neural network decides which robots should receive priority, the system employs a tried-and-true planning algorithm to tell each robot how to move from one point to another. This efficient algorithm helps the robots react quickly in the changing warehouse environment.

This combination of methods is key.

“This hybrid approach builds on my group’s work on how to achieve the best of both worlds between machine learning and classical optimisation methods. Pure machine-learning methods still struggle to solve complex optimisation problems, and yet it is extremely time- and labor-intensive for human experts to design effective methods. But together, using expert-designed methods the right way can tremendously simplify the machine learning task,” says Wu.

Overcoming complexity

Once the researchers trained the neural network, they tested the system in simulated warehouses that were different than those it had seen during training. Since industrial simulations were too inefficient for this complex problem, the researchers designed their own environments to mimic what happens in actual warehouses.

On average, their hybrid learning-based approach achieved 25 percent greater throughput than traditional algorithms as well as a random search method, in terms of number of packages delivered per robot. Their approach could also generate feasible robot path plans that overcame congestion caused by traditional methods.

“Especially when the density of robots in the warehouse goes up, the complexity scales exponentially, and these traditional methods quickly start to break down. In these environments, our method is much more efficient,” Zheng says.

While their system is still far away from real-world deployment, these demonstrations highlight the feasibility and benefits of using a machine learning-guided approach in warehouse automation.

In the future, the researchers want to include task assignments in the problem formulation, since determining which robot will complete each task impacts congestion. They also plan to scale up their system to larger warehouses with thousands of robots.

The research was funded by Symbotic.