Industrial goods industry trends 2027+

Summary: Copper hit successive records in 2026 while US Section 232 duties widened to conductors and cables. Makers of switchgear, busways, cables and HVAC equipment redesign toward aluminum where space allows, while compact motors and transformers stay on copper and pass price escalation to machine builders.

Current situation: LME copper set an intraday record of USD 14,527.50/t on 29 January 2026 and a new peak of USD 14,875/t on 10 September 2026. Copper concentrate treatment charges fell to minus USD 272/t in October as Escondida suspended operations after a fatal accident and Grasberg stayed below full rates. In June 2026 a US proclamation kept the 50% Section 232 tariff on semi-finished copper, extended it to electrical conductors and cables, and cut the domestic content threshold from 95% to 85%; refined cathode stayed exempt while Commerce reviewed the carve-out. The copper to aluminum price ratio hit a record 4.3 in January 2026, above the 3.5 to 4.0 range where firms consider switching.

Expected development: J.P. Morgan expects substitution to rise from about 2% of copper demand today to roughly 6% by 2030. By 2028, aluminum busbars, cable cores and HVAC coils become standard catalog options at switchgear, busway and cooling equipment makers, following Prysmian, which already runs about 40% aluminum by weight. Compact servo and IE5 motors keep copper windings, and their makers add metal escalation clauses. A US tariff on refined copper in 2027 would widen the US premium and speed North American redesigns.

Challenges: Aluminum conducts about 61% as well as copper and needs about 1.6 times the cross-section, which compact motors and transformers struggle to absorb. Redesign, retooling and requalification cost time and money, and aluminum carries its own tariff and war premiums in 2026. The counter-argument is that the market sits close to balance, with ING forecasting a 2026 deficit of about 35 kt, and that investor flows drove the January spike.

Time to impact: 0-2 years
Potential impact: High
STEEP segment: Economic
This year's trend development: Rewritten with a new thesis

Summary: Europe's flagship hydrogen steel projects were cancelled, delayed or switched to natural gas in 2025 and 2026, and buyers resist the premiums producers need. For machinery makers, low-carbon steel through 2030 means scrap-based EAF material and gas-based DRI, which sets the ceiling for product carbon footprint promises.

Current situation: ArcelorMittal scrapped its hydrogen DRI plans for Bremen and Eisenhüttenstadt in June 2025, forgoing EUR 1.3 billion of German subsidies, as hydrogen cost EUR 5 to 8/kg against the EUR 2.50 to 3.00 needed. In February 2026 it committed EUR 1.3 billion to a single 2 Mt EAF in Dunkirk for 2029. On 30 October 2026 thyssenkrupp received revised state aid terms that let its 2.5 Mt Duisburg DRI plant run on natural gas and add hydrogen gradually. In April 2026 spot demand for green steel was near zero, and buyers accepted flat steel premiums of EUR 100 to 150/t against the EUR 150 to 170/t producers say they need.

Expected development: Through 2030, low-carbon steel for machine frames, housings and plate comes from scrap-based EAF routes at about 0.6 t CO2 per tonne and from gas-based DRI, with hydrogen steel limited to Stegra's Boden plant and pilot volumes. Machinery makers will meet customer carbon footprint requests with EAF content and certificates, and certified flat steel premiums settle near the EUR 100/t buyers accept today. Duisburg follows hydrogen prices, which keeps full hydrogen operation beyond 2030.

Challenges: EAF routes depend on scrap supply and power prices, and Europe lacks harmonized green steel definitions, which keeps premiums opaque. Stegra's start-up timeline was under review in April 2026 despite EUR 1.4 billion of new financing. The strongest counter-argument is that CBAM reform and the EU's proposed steel tariff rate quotas raise the price of conventional imports enough to revive hydrogen projects after 2028.

Time to impact: 4-6 years
Potential impact: Medium
STEEP segment: Ecological
This year's trend development: Reframed as a counter-trend

Summary: Chinese suppliers hold 55% of China's industrial robot market and 85% in metal and machinery applications, and Chinese off-highway OEMs bring products to market in two years against three to five for Western peers. The assumed Western engineering edge erodes in volume segments, which puts price and margin pressure on European and Japanese OEMs.

Current situation: IFR data released in September 2026 show China installed 354,000 industrial robots in 2025, 59% of the global total, and Chinese suppliers installed 195,000 of them for a 55% home market share, up from about 30% in 2020. In metal and machinery applications, Chinese robot suppliers hold 85% of their domestic market. China's 15th Five-Year Plan for 2026 to 2030 puts AI-powered robotics at the core of industrial policy. In off-highway equipment, Interact Analysis reports that Chinese OEMs such as SANY, Liugong and Weichai develop products within two years against three to five for Western lines, run lower margins partly backed by subsidies, and build plants in Mexico, Brazil, South Africa and India.

Expected development: Between 2027 and 2030, Chinese suppliers extend from robots and wheel loaders into excavators, mining equipment and mid-range automation in Europe, Southeast Asia and Latin America. European and Japanese OEMs respond by compressing development cycles, sourcing Chinese components and locking customers into service and uptime contracts of five to seven years. Western share holds in high-precision, safety-certified and application-engineered segments, while price points in volume segments fall.

Challenges: Chinese suppliers' home share in robots dipped from 57% to 55% in 2025, so foreign suppliers still hold 45% of the largest robot market. Perceived quality gaps and tariffs slow adoption in Europe and North America, which is why Chinese OEMs localize production abroad. Counter-argument: Western firms retain leads in precision components and application know-how that unit share metrics underweight.

Time to impact: 0-2 years
Potential impact: Very high
STEEP segment: Economic
This year's trend development: New

Summary: PLC margins rest on proprietary hardware, I/O scaling and engineering environments. Virtual PLCs and agents that generate control code weaken all three, so automation vendors shift revenue toward licenses, services and recurring contracts, and industrial buyers gain leverage in control system tenders.

Current situation: ABI Research reported in May 2026 that virtual PLCs, offered only by Beckhoff and CODESYS until 2023, now also come from Siemens, Emerson and Schneider Electric, and it expects them to reach 16% of the PLC market by 2030, with Schneider holding 17% of the vPLC market. ABI argues that once I/O capacity depends on memory and licensing, and agentic AI writes control code, the engineering environment loses weight as a differentiator. Vendors already shift to recurring models: Siemens sells its Eigen Engineering Agent through a subscription, and Schneider trained more than 400 professionals from over 215 partner companies on its open automation platform between February and June 2026. Schneider cites Omdia data that closed automation environments cause an average of US$11.28 million in hidden annual costs.

Expected development: Between 2027 and 2031, the large PLC vendors move pricing from controller hardware toward runtime licenses, engineering subscriptions and lifecycle service tiers, and open their tools to customer AI agents through MCP or APIs. Price pressure hits mid-range controllers first, where IPC makers such as Advantech supply commodity compute. Vendors with SIL3 safety certification and deep application libraries defend margins, while vendors that rely on engineering-tool lock-in lose share in greenfield tenders.

Challenges: Installed-base inertia is enormous: ABI puts open software-defined automation at less than 1% of new PLC installations in 2025 and cites immature ecosystems and unclear support paths. Engineering software customers already voice concerns about lock-in and rising costs as Autodesk introduces usage-based charges for its APIs. Vendors also cannibalize hardware revenue before software income scales.

Time to impact: 4-6 years
Potential impact: High
STEEP segment: Economic
This year's trend development: New

Summary: NVIDIA plans 800 VDC distribution for megawatt racks starting in 2027 and has enlisted ABB, Eaton, Schneider Electric, Siemens, Vertiv and others. AI data centers now set specifications for DC protection, high-ratio conversion and rack-level storage, which pulls electrical equipment R&D budgets toward DC power electronics.

Current situation: In May 2025 NVIDIA announced a move to 800 VDC distribution for 1 MW IT racks, with full-scale production timed to its Kyber rack systems starting in 2027. It argues that 54 V rack distribution hits physical limits above about 200 kW, that a 1 MW rack at 54 V would need up to 200 kg of copper busbar, and that 800 VDC cuts copper by about 45% against 415 VAC and raises end-to-end efficiency by up to 5%. By October 2025 NVIDIA listed ABB, Eaton, GE Vernova, Heron Power, Hitachi Energy, Mitsubishi Electric, Schneider Electric, Siemens and Vertiv as data center power system partners, with Infineon, STMicroelectronics, Texas Instruments and others supplying silicon. NVIDIA and partners are evaluating solid-state transformer designs.

Expected development: From 2027 to 2029, electrical equipment makers in NVIDIA's ecosystem launch 800 VDC busways, solid-state breakers, sidecar power racks and supercapacitor storage for AI halls, and qualify solid-state transformers for grid-to-DC conversion. Products co-designed with chipmakers set the reference designs that hyperscalers buy, which favors vendors with power semiconductor expertise and global service networks. Factory DC microgrids adopt the same components after 2029.

Challenges: NVIDIA itself still weighs the cost, operating and safety implications of solid-state transformers against conventional designs, and solid-state protection devices for 800 VDC remain emerging. The timeline depends on one chip vendor's roadmap and on AI capital spending holding up, and a slowdown would strand R&D spent on DC portfolios.

Time to impact: 2-4 years
Potential impact: High
STEEP segment: Economic
This year's trend development: New

Summary: US manufacturers may leave 1.9 million of 3.8 million needed jobs unfilled by 2033, and German metal and electrical firms lack more than 121,000 skilled workers in core occupations. Retirements keep the gap open even in a weak cycle, so industrial goods makers plan robots, AI assistance and training together as capacity insurance.

Current situation: Deloitte and the Manufacturing Institute estimated in April 2024 that US manufacturing needs 3.8 million workers from 2024 to 2033, of whom 1.9 million could go unfilled. US manufacturing job openings rose to 584,000 in September 2026, a 4.4% rate, from 389,000 a year earlier, according to BLS. In Germany, KOFA counts a gap of more than 121,000 skilled workers in core metal and electrical occupations, with almost half of vacancies unfilled and the largest gaps in metalwork and in machinery and vehicle technology. IAB projects the potential workforce shrinking by 7.2 million by 2035 without migration at constant participation.

Expected development: From 2027 to 2030, the gap widens as baby boomers retire, even where employment falls in cyclical sectors such as machine tools. Manufacturers that pair automation with structured upskilling and AI-guided work instructions hold output with fewer hires, and firms that wait lose orders in welding, machining, maintenance and electrical trades. Labor availability becomes a plant siting criterion that favors locations near technical schools and with open immigration rules.

Challenges: Cyclical weakness masks the shortage: German machine tool employment fell 6% in the year to June 2026, and IFR describes Germany as in recession. Automation shifts demand toward scarcer skills such as robot programming and maintenance. Tighter US immigration rules and German migration politics could widen the gap beyond current forecasts.

Time to impact: 0-2 years
Potential impact: Very high
STEEP segment: Social
This year's trend development: Rewritten with a new thesis

Summary: About 7,000 full-size humanoids were sold worldwide in 2025, and carmakers piloting them typically run fewer than ten units. Announced fleets at Schaeffler and Hyundai start between late 2026 and 2028, so industrial goods makers should treat humanoids as an R&D option and keep automation capex on proven arms, AMRs and cells.

Current situation: The IFR counted about 7,000 full-size humanoids sold in 2025 and concluded in September 2026 that most remain prototypes or pilots, with industrial robots economically superior for production tasks. Carmakers piloting humanoids typically run fewer than ten machines, and many buyers use them to generate AI training data. BMW's Figure 02 pilot supported 30,000 X3 builds over ten months at Spartanburg, and in June 2026 BMW announced that Figure 03 will take on logistics sequencing. Schaeffler and Humanoid start an initial rollout at two German sites between December 2026 and June 2027, beginning with box handling in Herzogenaurach. Hyundai plans Atlas deployment at its Georgia Metaplant from 2028, starting with parts sequencing.

Expected development: Through 2027, humanoid fleets at any single plant stay below 100 units and concentrate on tote handling, kitting and parts sequencing in automotive and logistics. The real test of scale comes in 2028, when Hyundai's Metaplant rollout and Schaeffler's plan for 1,000 to 2,000 units by 2032 move past pilot stage. Suppliers of actuators, reducers, hands and sensors earn revenue first, as Schaeffler's seven-digit actuator deal with Humanoid shows. Wheeled mobile manipulators win most brownfield tasks that humanoid demos target today.

Challenges: Dexterity, autonomy outside controlled settings, the lack of a killer application and unresolved liability and insurance questions remain barriers, and Musk calls Optimus the hardest product to scale manufacturing that Tesla has made. The strongest counter-argument is Chinese volume production, where falling prices could make humanoids competitive with custom cells sooner.

Time to impact: 6-8 years
Potential impact: Medium
STEEP segment: Technological
This year's trend development: New counter-trend

Summary: Surveys from 2025 and 2026 agree that broad AI use has outrun measurable returns: MIT found 95% of businesses without measurable return from generative AI, and McKinsey counts only 6% of firms as AI high performers. For industrial goods makers this argues for fewer, deeper use cases tied to plant KPIs and for data infrastructure spending ahead of new pilots.

Current situation: MIT's Project NANDA reported in August 2025 that 95% of businesses saw no measurable return from generative AI and that only 5% of custom enterprise tools reached production. McKinsey's November 2025 survey of 1,993 respondents found 88% use AI in at least one function, about one third have begun scaling, 39% report any EBIT impact, mostly below 5%, and 6% qualify as high performers. Gartner predicted in June 2025 that over 40% of agentic AI projects will be canceled by the end of 2027 and counted only about 130 genuine agentic AI vendors. Rockwell's May 2026 manufacturing survey found 18% of firms still in pilot mode and only 43% of collected data used effectively.

Expected development: From 2027, manufacturers cut AI pilot portfolios and concentrate budgets on three to five scaled use cases such as visual inspection, maintenance triage and planning, each tied to scrap, uptime or inventory targets. Spending shifts toward data platforms, OT connectivity and process redesign. Vendors move to outcome-based pricing, and AI high performers among industrial goods makers widen their margin lead because scaled use cases compound across plants.

Challenges: Survey definitions of return differ, and MIT's 95% figure covers generative AI across all sectors, so it overstates failure for mature industrial AI such as inline inspection. Siemens and Procter & Gamble report 10% to 20% scrap reduction where AI sits inside a product. The gap could close faster as vendors package domain models and agents mature.

Time to impact: 0-2 years
Potential impact: High
STEEP segment: Economic
This year's trend development: New counter-trend

Summary: Data center, transmission and electrification orders outpace factory capacity for transformers, switchgear and HVDC equipment. Siemens Energy, ABB, GE Vernova and Hitachi Energy book far more than they ship and earn segment margins near 20%, and most of the capacity they add lands from 2028, which keeps pricing power with incumbents.

Current situation: Siemens Energy held a €162 billion order backlog in its fiscal Q3 2026, about €51 billion of it in Grid Technologies, with a book-to-bill of 1.57 and a 19.9% Grid Technologies margin. ABB's backlog reached $32.3 billion on 30 September 2026, up 35%, with Electrification orders up 70% and triple-digit growth in data center business. GE Vernova's Electrification backlog stood at $42.2 billion at a 19.1% margin. Hitachi Energy announced a $528 million transformer plant in Gallman, Mississippi, in September 2026 with production from 2029, plus about INR 2,000 crore for a Gujarat plant due in FY28, and ABB a $200 million European program in May 2026. The IEA sees annual grid investment needing to rise by about 50% from $400 billion by 2030.

Expected development: Transformer scarcity eases gradually from 2028 to 2029 as US, Indian and Canadian plants ramp, and pricing stays firm until then. Incumbents use the window to lock in multi-year frame agreements with utilities and hyperscalers. By 2028, Indian, Korean and Chinese suppliers win a growing share of distribution transformer and switchgear tenders in Europe and North America as buyers accept new vendors to secure delivery slots.

Challenges: Backlogs carry cost risk on copper, steel and labor, and execution problems can erase margins. Data center demand could slow if AI capex corrects, and transmission projects that take five to 15 years to plan and permit can push deliveries out. Capacity announced in 2025 and 2026 may overshoot once it lands around 2029.

Time to impact: 0-2 years
Potential impact: Very high
STEEP segment: Economic
This year's trend development: Rewritten with a new thesis

Summary: China now ships more machine tools than Germany and redirects export capacity toward Europe and emerging markets as US access tightens. German builders lose share in China and in third markets at the same time, which squeezes Mittelstand machinery makers in the mid-price segment where Chinese quality has caught up.

Current situation: VDW figures presented in February 2026 put China at 21.6% of global machine tool exports, ahead of Germany at 16.7% and Japan at 14.2%, with Chinese exports up 18% in a year and China producing 36.9% of the world's machine tools in 2025. VDMA reports German machinery exports to China fell 13.8% in the first half of 2026, while German machinery imports rose 2.2% to €47.9 billion. Germany's trade balance with China in high-end machinery and factory equipment swung from a $900 million surplus to a $600 million deficit in less than a year. Rhodium Group records Chinese import share gains of 3 points in Brazil and 7.2 points in Indonesia over three years, with machinery among the sectors where China moved ahead.

Expected development: Through 2028, Chinese builders of CNC machining centers, laser cutters, injection molding machines and standard automation gain share in Southeast Asia, Latin America and Eastern Europe, and German builders concentrate on customized systems, service and software. Expect EU anti-dumping or anti-subsidy cases on machinery categories by 2028, and Chinese buyers to take over more distressed German machine builders. Builders that bundle machines with software, retrofit and service contracts defend their margins best.

Challenges: German builders keep strong positions in customized lines, process know-how and installed-base service, which price competition reaches slowly. A yuan that has lost nearly 20% against the euro since mid-2022 drives much of the price gap and could reverse. EU trade defense on machinery risks Chinese retaliation against Germany's remaining €81.8 billion of goods exports to China.

Time to impact: 0-2 years
Potential impact: Very high
STEEP segment: Economic
This year's trend development: New

Summary: Reshoring announcements keep coming, yet US manufacturing construction spending sits about 32% below its September 2024 peak and factory employment remains below its 2023 level. Capital shifts to machinery and electrical equipment for existing sites, which favors suppliers with retrofit and automation offers over those betting on greenfield plant programs.

Current situation: Census data show manufacturing construction at a $169.8 billion annual rate in July 2026, about 32% below the $250.2 billion peak of September 2024 and down 22.2% year to date. Computer, electronic and electrical projects account for $27.7 billion of the $29.1 billion decline. Manufacturing employment fell by 89,000 in 2025 and stood at 12.6 million in June 2026, about 305,000 below the January 2023 peak. Equipment moves the other way: nondefense capital goods orders excluding aircraft reached $88.2 billion in September 2026, with year-to-date orders up 10.9%, and machinery unfilled orders rose 8.3% to $155 billion. The Reshoring Initiative still counted 247,000 announced reshoring and FDI jobs for 2025, matching 2024.

Expected development: Through 2028, US manufacturing construction settles well above the 2019 level of $81 billion and below $200 billion a year, while core capital goods spending keeps growing on data center, grid and defense demand. Suppliers of automation, drives and test equipment that sell brownfield upgrades capture most of the reshoring budget. Factory employment grows by tens of thousands per year, which keeps announced jobs far ahead of payroll gains.

Challenges: The Reshoring Initiative projects 309,000 announced jobs for 2026, and 36% of OEMs in its September 2026 survey say they reshore or are actively engaged in it, so a second wave of factory shells could start once tariff rules settle. Construction data also lag announcements by years, which makes any call on the bottom of the cycle uncertain.

Time to impact: 0-2 years
Potential impact: High
STEEP segment: Economic
This year's trend development: Reframed as a counter-trend

Summary: Machinery makers earn a disproportionate share of profit from parts, repairs and service contracts on their installed base. KSB's service segment produced a third of 2025 sales and almost three quarters of group EBIT, and Atlas Copco generates 39% of sales from service at higher margins than equipment. Capital allocation, M&A and R&D priorities follow that margin pool.

Current situation: KSB's 2025 results show its SupremeServ spare parts and service segment at €1,013 million of €3,035 million group sales and €184.5 million of €252.1 million group EBIT, while pumps earned €71.3 million and valves posted a loss. At its November 2025 Capital Markets Day, Atlas Copco put service at 39% of sales, with higher margins than equipment in all four business areas. Caterpillar reported $24 billion in 2025 services revenues and targets $30 billion by 2030, backed by digital tools and a growing installed base. Sandvik names aftermarket growth as a strategic priority for its Mining division, which lifted its adjusted EBITA margin to 20.7% in 2025.

Expected development: From 2027, more listed machinery makers report service as a separate segment with its own margin target, and Atlas Copco's service share moves above 40% by 2030. M&A shifts toward service networks, rebuild shops and multi-brand repair, and equipment pricing is set to lock in parts and service attach rates. Makers with low attach rates on their installed base trade at a valuation discount to service-heavy peers.

Challenges: Service growth depends on installed base growth and on parts pricing power, and both face pressure. Right to repair rules, the EU Data Act and independent service providers attack parts margins. Customers with in-house maintenance teams push back on contract prices as machines age, and Caterpillar needs until 2030 to move services revenue from $24 billion to its $30 billion goal.

Time to impact: 0-2 years
Potential impact: High
STEEP segment: Economic
This year's trend development: New

Summary: Since 12 September 2025 users of connected machines in the EU can access the data their equipment generates, free of charge, and route it to a repairer or service provider of their choice. From 12 September 2026 new products must expose that data by design. OEMs can block use for competing products, and aftermarket services are open to competition.

Current situation: The Data Act has applied since 12 September 2025, and its access-by-design duty covers connected products placed on the EU market from 12 September 2026. The Commission lists repair, maintenance and data-based insurance as aftermarket uses, gives a bulldozer as a worked example and bars use of the data to build a competing connected product. Agricultural machinery association CEMA reads diagnostic downloads with raw or pre-processed data as in scope and outputs of proprietary algorithms as protected. The Council's Digital Omnibus compromise text, reported in July 2026, lowers the threshold for trade secret refusals from highly likely to likely serious damage. Baker McKenzie found case law on such refusals still absent in October 2026.

Expected development: By 2028, independent service providers and multi-brand platforms in agriculture, construction and compressed air offer predictive maintenance on rival OEMs' machines using Data Act requests, and first national authority decisions on trade secret refusals define the limits. OEMs answer with standardized data portals and paid access tiers for third parties. Parts and service margins on machines older than five years come under pressure wherever independents gain diagnostic data.

Challenges: The Digital Omnibus may widen trade secret refusals before enforcement starts, and several member states still lack authorities. Shared data covers raw and pre-processed signals, so independents must build their own analytics. Customers exercise these rights mainly when an OEM relationship breaks down, and older machines lack the connectivity that triggers the duties.

Time to impact: 2-4 years
Potential impact: High
STEEP segment: Political
This year's trend development: New

Summary: Pay-per-use and outcome-based contracts remain a sliver of machinery revenue. Providers keep machines on their own balance sheets, price utilization risk they cannot control and rebuild billing and sales incentives, so the model stays confined to niches with easily metered output such as compressed air, printing and welding robots.

Current situation: IoT Analytics put equipment-as-a-service at under 1% of 2023 equipment sales and found only two of seven compressor makers selling true EaaS by 2024, at single-digit revenue shares for Kaeser and Atlas Copco. Carlsquare's December 2025 EaaS sector report shows sector growth of 0.9% in 2024, expects 2.8% in 2025 and 3.6% in 2026, then a 6% CAGR to 2032. The Open Industry 4.0 Alliance's April 2025 pay-per-use paper names the blockers: machines stay on the provider's balance sheet, pure usage fees miss the fixed costs of installation and servicing, revenue recognition and depreciation change, and providers carry utilization risk that depends on customer staff. Working cases cluster where output is easy to meter, such as Kaeser in compressed air, Heidelberg in printing and Path Robotics in welding.

Expected development: Through 2030, EaaS grows mainly as a financing wrapper: captive finance arms, banks and insurers carry the assets and utilization risk, and OEMs sell metered uptime on top. Hybrid contracts with a base fee plus a usage charge become the standard form. Pure outcome-based pricing stays below 5% of revenue at listed European machine builders through 2030, and compressed air, vacuum, printing and robot cells keep the lead because their output is measurable at the machine.

Challenges: The strongest counter-argument is customer pull: buyers under capex freezes ask for flexible terms, and Carlsquare reports valuation premiums for service and digital solution providers. If captive finance and insurers absorb balance sheet and usage risk at scale, adoption could accelerate faster than OEM capabilities suggest. Trust in metering data and channel conflict with dealers remain open.

Time to impact: 4-6 years
Potential impact: Low
STEEP segment: Economic
This year's trend development: Reframed as a counter-trend

Four shifts carry the most weight. US tariff law, with Section 232 duties on the full value of machinery and a pending decision on robotics and machinery imports, decides where machines for the US market get built. China has overtaken Germany in machine tool exports and supplies 55% of its own robot market. Rare earth, tungsten and gallium supply now depends on export licenses. EU product law, from the Machinery Regulation to the Product Liability Directive, turns machine software into a compliance and liability issue from late 2026.

Since 6 April 2026, Section 232 metal duties apply to the full customs value of listed derivative products, with a transitional 15% rate for listed industrial machinery, molds, dies and tools until 31 December 2027. Section 301 duties of 10% to 12.5% have applied to imports from more than 80 economies since 24 July 2026. A Section 232 proclamation of 28 October 2026 keeps tariffs on robotics and industrial machinery in reserve while Commerce and USTR negotiate for 180 days, with an update due in late April 2027.

The Machinery Regulation applies from 20 January 2027 and requires notified-body assessment for safety components with self-evolving behavior. Cyber Resilience Act reporting obligations have applied since 11 September 2026, with full obligations from December 2027. The Product Liability Directive covers software, AI and safety updates for products placed on the market from 9 December 2026. The AI Omnibus moved high-risk AI rules for machinery to 2 August 2028 and into machinery law. CBAM's definitive period, in force since January 2026, adds carbon costs and data duties for steel and aluminum inputs.

Exposure is high wherever designs use neodymium magnets with dysprosium or terbium. China's April 2025 licensing for seven medium and heavy rare earths and the magnets containing them stays in force, while the broader October 2025 package is paused until 10 January 2027. Pre-cleared civilian customers can buy under year-long general licenses from the largest Chinese magnet makers. Mitigation runs through non-China magnet plants backed by state price floors, magnet recycling and magnet-free IE5 synchronous reluctance motors for pumps, fans and compressors.

The evidence points to pilot scale through 2027. About 7,000 full-size humanoids were sold worldwide in 2025, and carmakers piloting them typically run fewer than ten units. Announced fleets at Schaeffler and Hyundai start between late 2026 and 2028. Volume automation spending in 2027 goes to robot arms, mobile robots and cells, and humanoids remain an R&D option.

Teams import the trend set into their ITONICS workspace, rate each trend with their own experts and link it to the projects, technologies and roadmaps it affects. Radar, matrix and portfolio views then show which programs carry exposure to a given trend, and AI-generated, cited insights summarize what is happening around each trend as evidence changes. The trend radar on this page runs on the same platform.