Energy now accounts for 23% of operating costs for businesses surveyed. More than half (54%) say rising energy prices pose a moderate or major threat to profitability, a figure that has grown with each successive EEM survey since they began in 2022. In sectors with the highest energy intensity, the exposure is considerably more acute. For these businesses, coping with a permanent energy crisis, managing energy use and managing financial performance have become the same task.

The response across industry has been decisive in intent. Nearly all organisations surveyed (98%) are already investing or actively planning to invest — up from 93% in 2024. Half are targeting Net Zero within five years. But financial discipline is shaping what gets funded and what gets deferred: 83% require a return on energy efficiency investment within five years, and 40% within two years. Nearly a third (31%) lack the specialist resource to implement projects. A further 29% report a digital skills gap, and almost a quarter (23%) say they do not have sufficient data to justify investment internally. The infrastructure for action is being built; the capacity to act on it is lagging behind.

"The cost of inaction is now harder to justify than the cost of investment, and businesses know it," said Mike Umiker, managing director of the Energy Efficiency Movement. "The share of organisations citing upfront cost as their primary barrier has fallen from 53% in 2024 to 43% today. That ten-point drop tells us something important: the financial argument for energy efficiency is landing. But progress is not accelerating to match. The barriers that remain are structural; a shortage of skills, a lack of specialist expertise, and in too many cases, insufficient data to make the internal business case. Those are not problems you can solve with a cheque."

The report also points to a significant broadening of where businesses are directing investment. The share of organisations prioritising transportation and logistics has risen from 37% in 2024 to 49% in 2026, reflecting a move away from individual asset upgrades toward a system-wide view of energy use. More than 60% have invested in energy audits and cloud data management to build the foundation for that wider approach.

"Energy efficiency is no longer a sustainability topic in isolation, it has become a test of industrial competitiveness and resilience," Umiker added. "81% of organisations say better financing or government incentives would increase their investment, and more than half say they need external support across four or more areas. The tools and the technologies exist. What is needed now is a stable framework across policy, finance, and industry to deploy them at the pace this moment demands."

The findings land at the centre of a growing global debate on industrial energy security. The IEA's Global Conference on Energy Efficiency convenes on 29-30 June, in Montreal, bringing together government ministers, industry leaders, and financial institutions to address exactly these challenges. EEM will utilise the report findings for the conference, connecting the voice of 2,000 senior business decision-makers directly to the policy and investment conversations underway.

The full report – Energy Efficiency Investment Report 2026: Rising energy costs are outpacing the efficiency response – covers major energy-intensive sectors across North America, Europe, Asia-Pacific, and Latin America. Survey respondents were senior decision-makers at small, medium and large organisations.

Download the full report here

Now, engineers at Princeton University have designed and built soft-rigid hybrid robots that move and shift without the need for motors or external pneumatic controls. To do this, the team combined a printable polymer, called a liquid crystal elastomer, with flexible electronics and folding techniques based on the art of origami. In an article published March 21 in the journal Advanced Functional Materials, a research team led by professors Emily Davidson and Glaucio Paulino reported that they used a 3D printer to create a reconfigurable soft robot that was able to repeatedly move without noticeable degradation.

As a demonstration, the researchers built a soft robot in the shape of a crane, a classic origami figure, that flaps its wings when powered with electricity. The crane moves without a motor. Instead, the robot’s motion relies on targeted heating in the polymer to control the wing flapping. The experiment also demonstrated that the Princeton soft robot can precisely and repeatedly move and return to its original shape without wear or distortion with real-time programmable sequences, a key feature for future applications.

Construction begins with a 3D printer and a special polymer

The system starts with a molten polymer printed into patterned zones using Davidson’s customised 3D printer. Despite its flexible nature, the polymer the researchers use is a liquid crystal elastomer, which means its internal molecules have an ordered structure. Davidson’s group are experts in controlling the structure of liquid crystal elastomers through molecular design, and controlling the nanostructure (in this case, the orientation) of polymers through printing, both of which were crucial to this project. The researchers programmed the printer to vary the internal orientation of the molecular structure of the polymer as it prints. Each of the patterned zones in the printed material features consistent molecular alignment. By stacking these zones and joining them in different ways, the researchers were able to create hinges in the material that bend in pre-programmed ways when the material is heated.

As part of the printing, the researchers also added flexible electronics into the hinges in the material. The printed circuit boards’ flexible structure allowed the researchers to embed them directly in the printing material rather than apply the circuits in a separate step. This simplifies fabrication and allows for greater consistency and functional integration of the circuit into the robot.

Davidson noted that a critical advance in the current work was the integration of 3D printed liquid crystal elastomers with printed circuit boards that could be commercially manufactured. The ability to co-design the liquid crystal elastomer hinges and the flexible printed circuit boards to drive actuation made the fabrication and control of these soft-rigid robots feasible.

Printed circuit boards control the heating that drives motion

Once embedded, these circuit boards allow the researchers to heat extremely specific areas of the polymer structure and perform closed-loop control using embedded temperature sensors. This heating takes advantage of the carefully structured polymer, causing the material to contract in ways that the engineers program into the polymer printing. These contractions trigger folding along hinges. To ensure the material folds only at the hinges, the researchers added light fiberglass panels to the flexible printed circuit boards in between the polymer hinges.

The researchers used mathematics derived from origami patterns to control the robots’ motion based on systems of folding and unfolding. Paulino’s research team has pioneered the use of origami to design medical implants, construction components and robotics. Recent projects include segmented soft robots that use origami systems to navigate complex paths, robots that can transform their shapes and adaptively reconfigure in a magnetic field, and programmable systems that can encrypt information and function as mechanical computers.

Paulino said an important feature of the design is that the software used to control the robot uses embedded temperature sensors in the origami to compensate for small errors that creep into the system as the robot repeatedly changes shape. Paulino said that the ability to correct these errors is key to soft robots’ durability.

The work began as an undergraduate thesis project

The development of the new robotic system began as David Bershadsky’s undergraduate thesis project at Princeton. Bershadsky, now in graduate school at the University of Texas-Austin, received his degree in electrical engineering in 2024 from Princeton.

Bershadsky said he has been interested in robotics since he pursued a project in high school developing swarm robotics that deployed individual robotic units that changed size. When he arrived at Princeton, he was interested in advancing the work.

“I was looking for a way to easily and repeatedly create unit cells that could transform based on volume,” he said.

Bershadsky thought that liquid crystal elastomers could serve as a vehicle to create that type of robotic system and approached Professor Davidson, an expert in the material, for advice. Davidson thought it was an intriguing idea.

“She said, ‘You should probably take Paulino’s origami engineering class to get a more formal background,’” Bershadsky said. One of the TAs in the class was Tuo Zhao, a postdoctoral researcher in Paulino’s lab who helped develop the research and is one of paper’s co-authors.

Researchers designed technology for manufacture

Bershadsky said the work is primarily an “integration of material science with robotics with a focus on manufacturability.”

“I think the big contribution is we showed integration of a complex system where we have local heating control,” he said. “We can control activation depending on where we heat.”

Besides demonstrating the viability of the robotic system, Bershadsky created a software tool that designers can use to create their own robots. The tool is available on the lab’s github and included with the paper’s dataset.

The most challenging part of the work?

“Honestly, people talk about system integration being the hard part,” Bershadsky said. “This entire project was the integration of a bunch of different technologies.”

engineering.princeton.edu

Honeywell's current brand value is estimated at $18 billion and has been synonymous with delivering innovation to customers for 140 years. Both new brands will draw on Honeywell's history while reflecting each company's focused strategy and long-term growth trajectory:

The automation business will be known as Honeywell Technologies and will continue to trade on the Nasdaq under the ticker "HON." The company will be a global leader of the industrial world's transition from automation to autonomy, with a comprehensive portfolio of mission-critical, outcome-based technologies, solutions and software to drive customers' productivity and growth.

The aerospace business will trade on the Nasdaq under the ticker "HONA" and will be known as Honeywell Aerospace. The company will be one of the largest publicly traded, pure-play aerospace suppliers, with leading positions in technology and systems that will continue delivering the future of aviation through increasing electrification and autonomous flight.

"Today marks another defining moment in our transformation into two independent, focused companies," said Vimal Kapur, chairman and CEO of Honeywell. "Drawing on Honeywell's century-long legacy, these new brand identities honor our history while reflecting the bold vision and strategic focus that will define Honeywell Technologies and Honeywell Aerospace as standalone companies. This is the start of an exciting new era for both businesses."

Honeywell Technologies

Honeywell Technologies' new brand demonstrates the company's clear focus and bold ambition. High-energy and kinetic, it retains the iconic Honeywell red as its anchor, alongside an expanded, more vibrant color palette and a modern "HT" monogram. The visual system, built around the concept of "intersections," reflects where the company's expertise meets technology to unlock precision, reliability and efficiency for its customers.

"Our new brand highlights the powerful intersections of our technology and expertise – from controls to intelligence to safety – that will redefine how industries operate, accelerating the shift toward a more autonomous future while unlocking new levels of growth and long-term value," added Kapur. "As Honeywell Technologies, this bold new chapter strengthens our foundation for the future and reflects our ability to continuously evolve our capabilities to deliver outcomes for our customers."

Honeywell Aerospace

The dynamic Honeywell Aerospace brand centers on a new, evolved logo – a stylised "H" and "A," with swooping negative space that suggests the line of the horizon. Its signature color is a distinctive sunrise orange, evoking the shades of the horizon as a pilot sees it at dawn, with silver accents that reflect the materiality of aircraft. This color palette gives Honeywell Aerospace its own distinct identity while still reflecting the trust that Honeywell has built over years.

"Our new Honeywell Aerospace brand reflects the precision, confidence and forward momentum that has defined the past century of innovation and trusted performance we have delivered for our customers and partners," said Jim Currier, president and CEO of Honeywell Aerospace. "As an independent company, we will be uniquely positioned to innovate faster, move with greater agility and shape the next era of aviation."

https://www.honeywell.com/gb/en

The investment arrives as global industrial operators face an urgent dual challenge: the need for productivity gains amid rising safety risks on factory and warehouse floors. Despite significant investment in automation, many facilities remain digitally dark regarding their manual fleets. According to the Occupational Safety and Health Administration, between 35,000 and 62,000 forklift-related injuries occur each year in the United States, resulting in an average of two fatalities every week. This safety risk exists alongside significant inefficiency, with forklifts productive for less than half of their total operating time. Despite heavy technology investment, most sites still lack real-time visibility into vehicle location and performance.

Slamcore has built the solution. Using a stereo camera and proprietary visual AI, Slamcore's technology continuously tracks the position and behavior of any vehicle in a facility without GPS, beacons, floor markers or any other infrastructure. Slamcore Aware gives operations managers facility-wide visibility of every vehicle, enabling smarter utilization, faster investigations and meaningful reductions in idle time. Slamcore Alert monitors driver behavior and proximity to pedestrians and structures, catching the near misses before they become incidents.

"Operations managers in factories and warehouses have largely been flying blind when it comes to their manual fleets. Slamcore Aware and Slamcore Alert change that from day one, without disruption to existing operations," said Owen Nicholson, CEO, Slamcore. "ROKStar Ventures' investment tells us that the industry's most sophisticated players see this as a foundational infrastructure, not just another point solution. As our footprint grows, so does a body of real-world operational data that does not exist anywhere else and that will become the backbone for the next generation of physical AI."

"Delivering visual AI that performs reliably at the scale and complexity of a real factory or distribution center is a genuinely hard problem," said Ryan Gariepy, vice president of Robotics at Rockwell Automation. "Most approaches either require significant infrastructure investment or fail to hold up in the dynamic, unpredictable conditions of an active facility. The potential for the same technology platform to work on every class of autonomous and human-operated industrial vehicle is key. We're also incredibly excited about their ability to scale without requiring complex and time-consuming vehicle or facility redesigns."

Jim Adler, founder and general partner at Toyota Ventures and a Slamcore board member since the company's earliest days, sees the long-term data opportunity as equally significant as the products themselves, stating: "At Toyota Ventures, we believe safety and efficiency go hand-in-hand. Slamcore Aware and Alert have proven this today, but their long-term potential is even more compelling. Each Slamcore deployment generates real-world operational data, which will train the next generation of physical AI models."

Operations teams looking to improve fleet visibility and reduce forklift incidents can learn more at:

slamcore.com

The project is being led by Redwire’s Luxembourg team, which recently completed several project milestones including preliminary design and performance assessment.

Before successful delivery, the MANUS Breadboard Model underwent a comprehensive test campaign to verify the functionality and performance of the manipulator and tool-changer subsystems, and to demonstrate operational scenarios aligned with system requirements. All planned operations, including payload handling, end-effector actuation with wireless data and power transfer, range extender manipulation, and automatic deployment, were executed successfully, confirming overall system readiness. Functional testing validated safe and reliable mechanical performance, demonstrating strong joint-space accuracy and stable interaction among subsystems.

MANUS is designed to support the loading and unloading of equipment and cargo from ESA’s Argonaut lander, which will enable Europe to deliver 1,500–1,800 kg of cargo, infrastructure, and scientific instruments to the lunar surface. The manipulator could facilitate loading and offloading to a rover, directly onto the lunar surface, or to an astronaut. MANUS could also function as a connector for rovers to transfer power and charge batteries, as well as a tool to scoop lunar regolith.

The Argonaut program aims to establish an enduring European presence on the Moon and ensure that Europe remains a leading participant in global space exploration.

rdw.com

Dr Sandra Schöttner, chief scientist for the Deep Arctic Expedition, Greenpeace International said: “This marks the deepest banner protest in history, to speak for ecosystems that have no voice of their own. World leaders have already promised to protect 30% of the oceans, now they must listen to the science and actually do it. We cannot meet our global goals if we also allow industrial exploitation of unexplored and vulnerable ecosystems in the deep sea. It is high time that leaders keep their promises and give the oceans a chance to recover.”

The banner was brought down to 2315 metres below sea level by the expedition’s remotely operated vehicle – ROV Holly – and held up in front of the hydrothermal vent field called “Loki’s Castle”, a unique volcanic ecosystem where black smokers are emitting 300–320°C fluid from deep within the ocean crust. Scientists believe our own distant ancestors may have looked just like the microbes living on structures like these, making Loki’s Castle a “cradle” of complex life that could hold the key to how life on Earth once started.

The Arctic is one of the most rapidly changing regions on Earth due to climate change, and as the industrial frontier expands toward the deep sea, through threats like deep-sea mining, Greenpeace warns that these unique ‘biodiversity hotspots’ are at risk of irreversible disruption.

Dr.Sandra Schöttner said: “It is not too late to act. Science clearly predicts the limits of what our planet can endure, but human action determines our survival. By safeguarding these deep-sea ecosystems within a global network of ocean sanctuaries and establishing a moratorium on deep sea mining, we can create a resilient safety net for marine life, and protect the health of our global oceans for generations to come.”

www.greenpeace.org

“For the BMW Group, the use of industrial data is a key factor in translating artificial intelligence into value creation,” said Dr. Franz Decker, CIO and senior vice president of the BMW Group. “By combining our engineering datasets with Mistral AI’s model training capabilities, we are building specialized AI which supports complex development tasks.”  

The scale and complexity of crash simulation at the BMW Group underlines the need for domain-specific AI. Each week, the company runs thousands of virtual crash simulations, generating vast amounts of engineering data. Over time, this has resulted in a historical dataset of over one petabyte of crash simulation data. It provides highly detailed insights into vehicle structures and material behaviour, forming a unique foundation for training an industrial AI model.  

“As Industrial AI becomes the new frontier for AI, we are proud to partner with the BMW Group” said Marjorie Janiewicz, chief revenue Officer of Mistral AI. “This collaboration shows how industry specific AI models can help solve complex engineering challenges such as crash simulation.” 

To scale this approach, the BMW Group is focusing on so-called Large Industry Models (LIM). These are AI systems trained on industry specific engineering and simulation data from vehicle development and safety testing. Unlike general‑purpose AI systems, LIMs embed domain‑specific knowledge directly into the AI model. This requires not only industrial data, but also deep domain expertise and technical environments that allow AI systems to learn directly from BMW’s development processes.  

The partnership highlights the importance of industrial data for the next phase of data‑driven value creation and strengthens the BMW Group’s AI and innovation ecosystem.

www.bmwgroup.com

The new Humanoid Robots – 2026 report from the market intelligence specialist predicts the long-term commercial inflection point will take place in 2032, with shipments exceeding 700,000 units in 2035 and market revenue reaching approximately $15 billion. However, this outlook remains conditional on achieving economic viability thresholds, as well as breakthroughs in embodied AI to enable autonomous, reliable task execution, clearer regulatory frameworks, and acceptable efficiency rates.

China and the US to dominate humanoid robot demand by 2035
 By 2035, Interact Analysis anticipates China will account for over 65% of real-world application shipments. This will be driven by government investment, subsidies, and state-owned enterprise procurement. The US market, in a distant second place, will see growth driven by capital markets, AI investment, and high labour costs. Together, China and the US will account for over 85% of demand for humanoid robots by 2035.

Short-term mass commercialisation of humanoid robots is restricted by immature core technologies and the lack of established regulations and industry standards. At present, industrial manufacturing and warehousing is leading near-term deployments due to structured environments and a high concentration of early techology adopters. This is followed by public services, driven by Chinese state-backed programs. Household use cases remain a longer-term opportunity, constrained by safety and environmental complexity.

Four end-use sectors forecast to experience transformative growth

Interact Analysis forecasts the following four end-use sectors will show significant growth in humanoid robot adoption through 2035:

• Real-world applications
• Academic R&D
• Robot training and data collection
• Entertainment.

While academic R&D and entertainment applications currently dominate production volume statistics, both are expected to grow at more moderate rates once markets mature. The robot training and data collection sector, on the other hand, is forecast to expand in the short term but stabilise over the long term as simulation technologies advance. However the most transformative growth is expected to come from the real-world applications sector, which is forecast to expand from around 10% of total production in 2025 to become the dominant market segment by 2035.

Marco Wang, Research Analyst at Interact Analysis, says: “Within the humanoid robots market, technology readiness remains a primary constraint, with gaps in embodied AI capability, severe data scarcity, and insufficient hardware durability and manufacturing consistency. Ecosystem and risk frameworks remain underdeveloped, with safety standards, certification pathways, and insurance mechanisms still required to enable economically viable deployment.

“The market is shifting from hype to pragmatism, with vendors and early adopters prioritising operational stability over headline specifications. For example, wheeled platforms are emerging as the preferred near-term form factor for real-world industrial deployment.”

www.InteractAnalysis.com

AEON, Hexagon Robotics’ multi-purpose humanoid, will be piloted on one of Fill’s client use cases in Gurten, Austria, to demonstrate how it can be integrated into real manufacturing workflows and as a complement to Fill’s automation solutions.

“We have focused on real-world applications from the outset, and it is encouraging to see strong interest from industry leaders such as Fill,” said Arnaud Robert, president of Hexagon Robotics. “This collaboration is an important step in validating how humanoids can increase the level of autonomy of the manufacturing processes and deliver measurable value across diverse environments. This includes improving efficiency on production lines, while expanding workforce productivity in complex manufacturing settings, such as high-mix production and precision assembly processes.”

With the pilot, Fill will deploy AEON across selected machine tending, inspection, and operational use cases, as a new way to combine Physical AI with its existing manufacturing solutions. The aim is to validate humanoids’ capability to enhance operational autonomy and flexibility, while generating value for customers.

“At Fill, we have a long tradition of continuously exploring innovative technologies that strengthen our offering,” said Andreas Fill, CEO of Fill Maschinenbau. “AEON's set of capabilities is unique, particularly its ability to operate in real industrial settings and integrate with existing systems. This collaboration allows us to test humanoid robotics in a structured and practical way, with a clear focus on delivering value to our customers.”

AEON is designed to operate alongside the industrial workforce, supporting tasks across manipulation, inspection, and data capture. This is made possible by AEON’s unique sensor fusion capabilities, spatial intelligence, and wheel-base locomotion.

The pilot will provide both companies with valuable insights into deployment, performance, and integration in real manufacturing settings. The collaboration reflects a shared commitment to advancing innovation in manufacturing and accelerating the adoption of next-generation automation solutions.

robotics.hexagon.com

Atlas Copco’s new Common Control Platform helps address these challenges by providing a unified, future-ready architecture. The value of a single interface enables manufacturers to operate multiple Atlas Copco technologies through a single control environment. For automotive OEMs managing hundreds of tools across a line, this translates into measurable gains in productivity, quality, and operational agility.

Fully compliant with the EU Machinery Regulations 2023/1230 which come into force in 2027, the platform provides a modular security framework that streamlines approvals and certifications.

The Common Control Platform has been developed to integrate Atlas Copco’s 8000 series of automated joining solutions and all applications:

•     Self-pierce riveting
•     HyRev fastening
•     Adhesive dispensing, potting and bonding 
•     Robot guidance
•     Metrology solutions, including laser vision inspection and quality assurance.

Built on a secure standardised interface and a unified HMI, the platform differentiates each technology’s specific features whilst sharing core architecture, whether in high-volume automotive production or the precision-driven assembly of power electronics components.

By consolidating disparate systems into one interface, the platform provides lower CAPEX through common hardware, eliminates traditional silos and enables faster commissioning, lowers TCO via simplified maintenance, and reduces training requirements.

The platform directly addresses core production challenges by enabling faster line integration through a unified control system that accelerates the deployment of new tools and line modifications while reducing training requirements. It also improves uptime by providing centralised monitoring that supports predictive maintenance and faster fault identification, helping to minimise costly downtime.

In addition, the platform enhances operational consistency through standardised control logic delivered via a common motherboard, ensuring stability across production stations and supporting quality assurance in high-volume automotive manufacturing environments.

Unified HMI

Utilising a web-based Human Machine Interface (HMI), the Common Control Platform provides users with a consistent experience across Atlas Copco joining technologies. The HMI features light and dark modes, offers intuitive navigation and requires no installation.  

As a unified interface, the platform reduces training time and  minimises user errors. By providing familiar and streamlined interface, the platform empowers operators to work more efficiently and confidently.

Link to data driven services

A Common Data Model provides a unified data structure which enables seamless integration with MES systems, data lakes or customer applications.  Atlas Copco’s ACDC Unify solution offers a single access point, application programming interface (API) and a normalised data model for all controllers.  This results in reduced engineering effort by avoiding integration errors and ensuring consistent, high-quality production data.  

By standardising and structuring data at source, the solution also creates a future ready foundation for advanced analytics, including the integration of artificial intelligence and machine learning applications as part of the overall production environment.

Manufacturers benefit from simplified connectivity and the use of standard interfaces such as OPC UA and MQTT.  This solution provides a scalable foundation for traceability, monitoring and analytics, such as Atlas Copco’s ALTURE system.

Cybersecurity focus

Based on IEC 62443, the international series of standards for the cybersecurity of Industrial Automation and Control Systems (IACS), the platform provides a modular security framework that streamlines approvals and certifications:

• Complies with the EU Machinery Regulation 2023/1230 by 2027
• Supports the EU-wide NIS2 Directive (EU 2022/2555) for cybersecurity
• Supports the Cyber Resilience Act (CRA) and the Radio Equipment Directive (RED) through a secure-by-design approach and a standardised software stack.

Overall, the platform delivers secure communication protocols across devices, with centralised access control and user management to reduce the surface for cyber threats.

This approach aligns with the broader shift toward secure-by-design industrial automation, helping OEMs protect both production continuity and sensitive manufacturing data.

Atlas Copco equipment already demonstrates conformity with these directives through CE and UKCA compliance processes, ensuring alignment with essential health and safety requirements. 

www.atlascopco.com