Reinventing Industry: Industry x.0 and Materials Acceleration Platforms
How and why materials and manufacturing startups are shifting to the full-stack model
Technology in the Physical World
The concept of ‘industrial technology’ or ‘hard tech’ is inherently far-reaching and difficult to pin down. Your mind may jump to CAT tractors, SpaceX rockets, or factory robots and you’d be right—the space is broad and only expanding as technology capabilities grow more multimodal and digitization of legacy industries becomes the norm and not the exception.
In its simplest form, industrial technology simply refers to the application of technology to physical processes such as manufacturing and energy generation. As a result, the space is unique as compared to traditional technology verticals as entire subsectors can spring into existence overnight as a result of material, often academia-driven, innovations. This can be contrasted with technology verticals where step-changes in technology stem from reorganizing existing key processes within an industry or workflow (ex: MDR pricing leading to SMB moat for Stripe and Shopify).
Taking industrial technology as the tech that interacts with the physical world either directly such as CNC machines and other hardware or through integration with the built environment such as biometric sensors, some important subsectors within the space to frame your understanding are:
Industrial design and simulation software
Computer Aided Design (CAD)
Digital Twin Software
Computer Aided Engineering (CAE) / Simulation Software
Manufacturing automation software
Product Lifecycle Management (PLM)
Management Execution Software (MES_
Energy management software
Building Energy Management Systems (BEMS)
Industrial Energy Management Systems (IEMS)
Residential Energy Management Systems (REMS)
Utility Energy Management Systems (UEMS)
Advanced Distribution Management Systems (ADMS)
Digital Fabrication Software
CNC Routing Software
3D Printer Planning Software
Supply chain software
IoT
These spaces encompass some of the largest technology companies globally. Some notable names that are market leaders in industrial technology include Palantir, Flexport, and Xometry and some of the most exciting technology verticals today such as supply chain and logistics (Amazon opens a 3 million square foot robot fulfillment center, stocking over 32 million items)
Taking the example of auto manufacturing, you can begin to envision how the technology involved in the manufacturing process has revolutionized the industry since the 1800s and in smaller revolutions since then such as robotic assembly (Devol and Engelberger’s Ultimate) and ultra-high-strength steel.
The combination of mass-production machinery, industrial software, IoT, and the layers of orchestration and collaboration software over automated production is understood as the 4th Industrial Revolution, or Industry 4.0
As you may have noticed with the example of auto manufacturing, the abstraction from the simple integration of conveyor belts in the assembly line to IoT points to modern revolutions in industrial technology being far more heterogenous in their applications—whereas most factories employ some sort of robotic automation, IoT-powered sensing differs across every application. This is partly why no one seems to agree what the eras of industrial technology are, however universally the current leap in manufacturing technology marks a generational shift: from machines aiding humans to humans aiding machines.
The most interesting startups defining Industry 5.0 today are seen making this leap, accelerating the flywheel of innovation and production using technology and ‘wrapping’ this in human oversight to mimic the order-of-magnitude boosts in production seen in previous industrial revolutions. This is often described as ‘full-stack’ or ‘in-a-box’ startups where disjoint tech-enabled processes are connected to automate an entire production, manufacturing, or research process as opposed to targeting independent workflows within a vertical. Companies like Helixx are applying the “in-a-box” model to outcompete traditional manufacturing through an automation-first platform while benefiting from the scaling of technology vendors as the company and similar startups primarily drive growth from low-cost adoption and scale.
This shift is not unique to industrial technology either, as vertical SaaS has been experiencing a similar ‘unbundling’ with a move away from core systems of record to a combination of function-specific winners driven by an automation-focused orchestration layer. This is exemplified by companies like Stord, optimizing logistics and fulfillment for e-commerce brands, which spun out order management and fulfillment from giant ecommerce platforms (Shopify) as an independent offering which would be orchestrated with other ecommerce software (Inventory management, returns management, content management, marketing) through overarching orchestration software (horizontal integrations: Zapier, Trayio. Vertical integrations: Shogun, Chord Commerce, Commercetools)
What does this mean for the future of industrial technology? These “in-a-box” platforms will expand into the physical world. Industrial technology is unique in this sense that value can be captured and moats can be developed by developing a physical execution wedge such as robotics or omnichannel data and sensing.
Case study: Zetwerk Mafia
To better illustrate this point, I want to highlight some companies that are employing this full-stack model to capture value as process-oriented businesses and as technology vendors.
Zetwerk is a leading example of this model based in India. The company has experienced meteoric growth, reaching a $2.7 billion valuation (CITE THIS), $580+ million in funding raised, and $100+ million in ARR. The company provides a marketplace that connects manufacturers to buyers such as OEMs (Original Equipment Manufacturers) and EPC (Engineering Procurement Construction). Zetwerk allows OEMs and EPC companies to manage partnerships with suppliers in 15+ countries and manage the entire design, manufacturing, and fulfillment process. The platform is typically used for buyers to gain broader access to precision-manufactured parts such as die-cast, CNC-machined, and injection-molded components in Zetwerk’s focus markets such as India, Vietnam, and Mexico.
Starting in 2025, Zetwerk has begun shifting focus to its in-house production capabilities to streamline the order-to-quote process and cater to more specialized and turnkey user demands, specifically for electronics and testing equipment such as test-rigs and process control tools. This comes as a long-term shift for the firm as they began in-housing QA in mid-2022.
Similar to notable Western talent ‘mafias’ from companies like Palantir and PayPal, Zetwerk has curated an entrepreneurial and high-agency culture that has led to many top executives exiting to launch their own manufacturing-focused technology businesses. Scimplify, CBCatalyst, and Sanlayan, all founded by ex-Zetwerk executives, have raised millions in 2025 backed by notable investors such as Accel and Lightspeed. These companies employ an end-to-end industrial technology model where they sell software for their target industries and leverage this software to gain a competitive edge and ‘double-dip’ on this value.
On a larger scale, many interesting companies are employing this approach within this industrial technology with some notable names being:
Nimble: Runs fully automated fulfillment centers using AI-powered robotics to handle picking, packing, and shipping for e-commerce brands.
Zetwerk: A global manufacturing platform that connects OEMs and EPCs with suppliers for precision parts, now expanding into in-house production.
Saronic Technologies: Builds autonomous surface vessels for naval operations with integrated software, sensors, and in-house manufacturing at Port Alpha.
Epirus: Develops high-power microwave systems like Leonidas that can disable drones using software-defined targeting.
Solugen: Produces industrial chemicals using enzyme-driven, fossil-free processes inside modular carbon-negative factories.
Furno: Makes compact, modular kilns to decarbonize cement production with localized, low-energy processing.
True Anomaly: Designs AI-guided satellites and mission software for space defense and in-orbit awareness.
Vatn Systems: Builds low-cost autonomous underwater vehicles for swarm-based defense and industrial missions
CBCatalyst: A full-stack R&D platform that automates chemical synthesis and materials testing using robotics and AI.
Materials Acceleration Platforms
The advanced materials (AdMats) industry is currently one of the most exciting spaces within industrial technology and manufacturing. The abundance of capital for industry-ready deeptech has allowed materials startups to reach commercial viability more quickly than ever. Take metallurgy, the study of the physical and chemical behavior of metallic elements and their alloys,, that is revolutionizing mass manufacturing and cutting millions in costs from global supply chains: Boeing is using 3D printed titanium parts in Apache helicopters, self-healing metals are extending the lifetime of oil refineries and turbine-blade engines, and revamped Iron-equipped battery designs in Inlyte Energy’s sodium-metal-halide batteries are leading to 2x longer lasting and cost-effective batteries as compared to traditional lithium ion designs.
To understand how such companies fit into the larger materials technology landscape, it’s important to consider the dynamics of how revolutionary AdMats are researched, discovered, produced, and applied in broader usage. The structure of the AdMats industry can be understood as
Discovery
This stage is driven by AI-powered platforms that simulate, predict, and optimize new material properties before anything is physically produced. Tools like Citrine, Albert Invent, and Orbital Materials enable inverse design workflows—starting from target performance specs and working backward to generate viable candidates. This phase is foundational to the concept of “self-driving labs,” as explored in Words Worth Doing, Kompas VC, and Battery 2030+.
Production
Once promising materials are discovered, startups use in-house pilot plants or synthesis labs to produce and test them under real-world conditions. Companies like InventWood (super-strong wood alternatives), Sila Nanotechnologies (next-gen silicon battery anodes), Phoenix Tailings (rare earth metals with zero waste), and Cyclic Materials (rare earth recycling) are building out production infrastructure to scale their materials safely and reliably.
Distribution
At this stage, procurement platforms like SAP Ariba or Coupa manage sourcing, but they’re not purpose-built for materials—treating advanced materials like any other commodity. This misalignment creates friction in high-complexity workflows and slows down adoption for R&D-intensive industries.
Processing
Marketplaces like Xometry serve as infrastructure connecting material vendors, processors, and manufacturers—offering quoting, routing, and execution across CNC machining, 3D printing, and finishing. These platforms compress fragmented industrial networks into API-accessible nodes, enabling seamless execution from material input to finished part. More context in Words Worth Doing: Manufacturing Startups.
What these breakthroughs and the companies behind them have in common is that they have bridged the gap between discovery, production, and distribution. Making this process repeatable, that is accelerating the discovery -> distribution process to BE the business model rather than the distribution itself, is the central value proposition of Industry 5.0 machinery through robotics, AI-driven research, and in-house production.
The concept of a materials acceleration platform (MAP) refers to this rolled-up materials supply chain and encompasses a research lab and production facility where research plans are executed, iterated, and re-executed to lead to production-ready material innovations, all powered by AI to develop research and robotics to automate production. MAPs act as autonomous labs where experiments are guided and evaluated using computational methods and human supervision and industry-ready manufacturing is integrated in a closed-loop system. This approach is interesting as it leads to orders-of-magnitude improvements to AdMats R&D processes while cutting labor costs involved in existing research labs.
MAPs are already being used to drive discovery in batteries (BIGMAP), biological tissue engineering (MAPs@Fraunhofer ISC, 44% decrease in labor required for nanoparticle synthesis), and development of high-entropy alloys (SOLID-MAP, 10-50x faster development). This model is behind some of the most interesting full-stack hard-tech startups today. A few to keep an eye on are: (shoutout to Justin Lopas at Worth Overdoing):
Osium AI: Uses AI to predict material properties and design sustainable formulations for chemicals and advanced materials.
Orbital Materials: Applies generative AI and simulation to rapidly design climate-focused materials like CO₂ sorbents.
Periodic Labs: Backed by ex-OpenAI researchers, aiming to build a “ChatGPT for materials science” with new startup funding.
Trace.Space: Offers an AI-powered requirements platform for engineers, recently raising €3.8M to halve R&D development cycles.
Dunia Innovations: Berlin-based startup with $11.5M funding to build self-driving material labs for green energy applications.
C1 (Circular Carbon): Developing quantum-chemistry-designed green methanol catalysis to decarbonize shipping and industry, recently backed with €20M.
Chemify: Launched a fully automated “Chemifarm” in Glasgow to turn chemical code into physical molecules with AI and robotics.
Layup Parts: Raised $9M (led by Founders Fund) to automate composite part manufacturing using robotics and integrated workflows.
Re:Build: Raised ~$600M (including $120M from General Catalyst) to create a vertically integrated U.S. manufacturing ecosystem.
Machina Labs: Uses AI and robotics to automate sheet metal forming, offering on-demand, mold-free manufacturing for aerospace and defense.
Saeki: Developing robotic microfactories to automate additive and subtractive manufacturing for high-performance ceramic parts.
Manukai: Builds an AI copilot for CNC machines that learns from past jobs to automate high-precision metal part programming.
Modern Synthesis: Provides AI-first software and robotics to automate discovery and scale-up of novel molecules and materials.
Note that these startups are at different points of the pure software - pure manufacturing spectrum but they illustrate the move from physics AI to physical AI, more here for the differences
I think the move from traditional software offerings in manufacturing such as manufacturing ERPs towards a more integration with production is undeniable. The ability to automate research to production pipelines is growing unavoidable for today’s manufacturing startups. Materials startups are poised to uniquely benefit from this as the medium is singularly driven by research which can be truly automated for the first time ever with MAPs. My questions would be whether value lies in “better” manufacturing itself, or selling the infrastructure to automate this manufacturing, MaaS or “in-a-box” startups like Helixx.