5 Composites Startups at JEC 2026 That Could Change Manufacturing Forever

From Nature-Inspired Fiber Placement to Shape-Shifting Molds — The JEC Startup Booster Class of 2026 Is Rewriting the Rules of Composites Manufacturing

The JEC Startup Booster — The Composites Industry's Premier Launchpad

Before diving into individual startups, it is worth understanding what makes the Startup Booster special. This is not a generic tech accelerator. It is the only global competition focused exclusively on composites and advanced materials innovation — and its track record of picking winners is remarkable.

By the Numbers

Table 1: JEC Startup Booster program statistics [1][2].

2026 Innovation Fields

The 2026 competition organized finalists across five fields of innovation [1]:

2026 Schedule

Table 2: JEC Startup Booster 2026 pitch and awards schedule [1].

The jury includes Jelle Bloemhof (Head of Composite Manufacturing Technologies, Airbus), Dr. Chris Skinner (VP Strategic Marketing & R&D, Owens Corning), Aline Rotzetter (Emerald Technology Ventures), Mike Salciccioli (ExxonMobil/Proxxima), and Uday Vaidya (CTO, IACMI) [1]. When Airbus's head of composite manufacturing is personally evaluating your startup, you know the stakes are real.

The Alumni Effect: What Past Winners Have Achieved

The Startup Booster is not just a pitch competition — it is a launchpad. The trajectory of past winners demonstrates why winning (or even being a finalist) at JEC matters.

All 20 Finalists at a Glance

Before spotlighting the six most disruptive, here is the complete 2026 class:

Table 3: All 20 JEC Startup Booster 2026 finalists [1][2].

What stands out immediately: six of the twenty finalists are in the digital/AI/software category. The composites industry — long dominated by materials science and mechanical engineering — is undergoing a digital transformation that is now spawning its own startup ecosystem. The raw materials category is equally strong, with seven companies proposing alternatives to conventional fibers and resins.

Startup #1 — Fibionic: The World's Fastest Fiber Placement

The Problem

Traditional automated fiber placement is precise and powerful — but it is fundamentally a serial process. An AFP head lays one tow or tape at a time, building up a laminate course by course. For aerospace structures where every ply matters, this is the right trade-off. But for high-volume automotive, consumer goods, and industrial applications where cycle times must be measured in minutes rather than hours, conventional AFP cannot compete with metal stamping or injection molding on speed.

The Solution

Fibionic's founders — Johannes Mandler, Thomas Rettenwander, and Elias Hirschbichler (who joined as co-founder in 2024) — took inspiration from nature. Biological load-bearing structures like bones and wood grain do not place fibers sequentially; they grow them simultaneously along stress paths. Fibionic's proprietary FFP (Fibionic Fiber Placement) process uses an airstream fiber deposition method that builds entire fiber layers in milliseconds — not seconds or minutes, but milliseconds [9][10].

The result is a reinforcement preform — a "skeleton" of fibers placed precisely along load paths — that can then be overmolded via injection molding, compression molding, or other conventional processes. Because fibers are placed only where structurally needed, the process produces zero fiber waste.

The Numbers

Table 4: Fibionic FFP vs. conventional AFP comparison [9][10].

The first fully automated production machine is now operational in Innsbruck, and it uses commingled yarns — hybrid yarns where reinforcing fibers (carbon, glass, or natural fibers) are intimately blended with thermoplastic matrix fibers (PA6, PA11, PA12, PP, or PET). When the preform is subsequently heated and pressed, the thermoplastic fibers melt and consolidate around the reinforcing fibers.

Why This Matters

Fibionic is not competing with AFP for aerospace structures. It is competing with metal stamping and injection molding for high-volume structural parts. The 500,000-parts-per-year capacity per machine places FFP squarely in automotive production territory. If the technology delivers on its promises at scale, it could open composites to markets where cycle time has always been the disqualifier.

Figure 1: Fibionic FFP Process Schematic

Airstream fiber deposition creating load-path-optimized reinforcement preforms, followed by overmolding via injection or compression molding.

Source: Fibionic [9][10]

Startup #2 — CGreen: Bio-Based Carbon Fiber from Recycled Cellulose

The Problem

Carbon fiber is one of the most energy-intensive materials on Earth. Conventional production starts with polyacrylonitrile (PAN) — a fossil-fuel-derived polymer — which is oxidized, carbonized at 1,000–1,500°C, and surface-treated. The process consumes approximately 230–300 MJ/kg and generates a substantial carbon footprint. PAN accounts for roughly 50% of carbon fiber production cost, and over 90% of global PAN precursor capacity is concentrated in Japan, the United States, and China [11].

For an industry that sells its products on the promise of lightweighting for sustainability, the carbon footprint of carbon fiber itself is an uncomfortable contradiction.

The Solution

CGreen replaces fossil PAN with cellulose sourced from recycled paper and recycled cotton [12][13]. The company is the product of ten years of R&D conducted within the FORCE project, led by IRT Jules Verne and supported by Forvia (formerly Faurecia). The proprietary spinning and carbonization processes are designed to produce carbon fibers with mechanical properties competitive with standard-modulus PAN-based fibers while cutting the CO₂ footprint by a factor of three [12].

Milestones and Backing

• €2 million raised in initial funding [13]
• Pilot carbonization line to be deployed at Icam (Carquefou, France) — a long-standing partner in the FORCE project [13]
• Selected by Dassault Systèmes' 3DEXPERIENCE Lab for industrial startup acceleration [12]
• Industrial validation of fiber properties targeted by 2027 [2]

The Significance

Table 5: Conventional vs. CGreen bio-based carbon fiber comparison [12][13].

CGreen is not the first to explore cellulose-based carbon fiber — the concept dates to Edison's carbonized bamboo lamp filaments. But previous cellulose-based fibers suffered from inferior tensile strength compared to PAN. What makes CGreen credible is the decade of systematic R&D behind it, the backing of IRT Jules Verne (France's premier advanced manufacturing research institute), and the use of recycled rather than virgin cellulose.

Figure 2: Carbon Fiber Precursor Pathways

Comparison of conventional PAN-based and CGreen cellulose-based carbon fiber production routes.

Source: CGreen [12][13]

Startup #3 — Fyous: The Shape-Shifting Mold

The Problem

Tooling is the composites industry's dirty secret. For every unique composite part geometry, a dedicated mold must be designed, machined (typically from Invar, steel, or aluminum), validated, and maintained. Tooling costs represent 30–60% of total composite part cost for low-to-medium volume production. Lead times for aerospace-quality tools run 8–16 weeks. And when a design iteration changes the part geometry by even a few millimeters, the tool often needs to be remade.

This is the single biggest reason composites cannot compete with metals for low-volume, high-mix production: the economics only work when the tooling cost is amortized over thousands of identical parts.

The Solution

Fyous has built what they call PolyMorphic Moulding — a mold made of more than 28,000 densely packed, individually addressable pins that can be digitally reconfigured into any freeform surface [14][15][16]. A flexible interpolation membrane sits over the pin array to create a smooth mold surface. The entire system reconfigures in under 20 minutes.

The Performance

Table 6: Traditional tooling vs. Fyous PolyMorphic Moulding comparison [14][15][16].

The first commercial system, the PM-01, supports molding, casting, vacuum forming, and composite layup processes [15]. Fyous is targeting applications across footwear, orthotics, medical devices, aerospace prototyping, and architectural panels.

Why This Matters for Composites

The implications for composite manufacturing are profound. Consider the aerospace MRO (maintenance, repair, overhaul) sector: every composite repair patch requires a caul plate or mold matched to the damaged area's geometry. Today, these are either machined (expensive, slow) or approximated (compromising repair quality). A PolyMorphic mold could be digitally configured to match the exact repair geometry in minutes.

For R&D and prototyping, the calculus changes entirely. Instead of committing $50K to a tool before validating a design, engineers can iterate through dozens of geometries on the same machine in a single day.

Figure 3: PolyMorphic Moulding Concept

Fyous pin-based reconfigurable mold system showing digital control, pin array, and flexible membrane.

Source: Fyous [14][15][16]

Startup #4 — Novium: Generative AI for Composites R&D

The Problem

Composites R&D is painfully slow. Qualifying a new material system for aerospace can take 5–10 years and millions of dollars. The industry generates enormous volumes of test data — coupon-level mechanical properties, process parameters, failure modes, environmental aging — but this knowledge is fragmented across PDF reports, proprietary databases, tribal expertise, and decades of published literature. When an engineer needs to know whether a specific resin system has been tested at a specific temperature with a specific fiber architecture, the answer is usually buried somewhere — but finding it requires hours of manual searching.

The result is a chronic cycle of reinvention. Teams unknowingly repeat experiments, miss relevant prior art, and default to conservative designs because the cost of exploring alternatives is too high.

The Solution

Novium is building a generative AI platform that transforms fragmented knowledge into structured intelligence [17][18]. The platform enables R&D teams to explore possibilities, anticipate outcomes, and make confident decisions before committing to physical testing.

While specific technical architecture details are limited at this stage, the value proposition targets the core bottleneck: converting the composites industry's vast but disorganized knowledge base into an actionable decision-support system. This means not just search (finding relevant data) but synthesis (connecting data points across sources to generate insights) and generation (proposing experimental designs, material combinations, or process parameters that optimize for multiple objectives simultaneously).

The Context

Novium is riding a wave of AI adoption in materials science. Companies like Materials Zone, Citrine Informatics, and Uncountable have demonstrated that machine learning can accelerate materials development cycles by 2–5× in adjacent industries (polymers, alloys, coatings). But composites present unique challenges: the design space is enormous (fiber type × resin × architecture × process × environment), the data is highly heterogeneous, and the physics of failure is anisotropic and scale-dependent.

A generative AI platform purpose-built for composites — rather than adapted from metals or pharmaceuticals — could unlock significant R&D productivity gains across the industry.

Why This Matters

Startup #5 — P2M / GENE.SYS: The Digital DNA of Composite Products

The Problem

A composite part has no memory. Once it leaves the factory floor, its manufacturing history — cure cycle parameters, layup sequence, resin batch, fiber lot, operator, ambient conditions — exists only in paper travelers or disconnected databases. During its service life, maintenance events, repairs, and loading history are tracked (if at all) in separate systems. At end of life, a recycler picking up a composite panel cannot tell what resin system it contains, what fiber type, or whether it has been contaminated — making intelligent recycling impossible.

This information asymmetry is about to become a regulatory problem. The EU's Ecodesign for Sustainable Products Regulation (ESPR) is mandating Digital Product Passports (DPPs) for a growing list of product categories. Composites — complex multi-material structures with long service lives and significant end-of-life challenges — are exactly the kind of products that DPPs are designed for.

The Solution

P2M's GENE.SYS system embeds a chip directly into the composite part during manufacturing, creating what the company calls the "digital DNA" of the product [19][20]. This chip carries a unique identifier linked to a cloud database that tracks the part across its entire lifecycle:

The technology is already being deployed commercially. Italia Yachts has become the first shipyard to integrate GENE.SYS, using it to track composite yacht components from production through maintenance [20].

Why This Matters

GENE.SYS sits at the intersection of three converging forces:

Figure 4: GENE.SYS Lifecycle Tracking

Composite part lifecycle with embedded GENE.SYS chip tracking from manufacture to end-of-life.

Source: P2M [19][20]

Startup #6 — Plastalyst (AC Biode): Recycling the "Unrecyclable" at Low Temperature

The Problem

Thermoset composites are, by chemistry, permanent. Epoxy, polyester, and vinyl ester resins cure through irreversible cross-linking — once set, they cannot be melted, reshaped, or dissolved under normal conditions. This makes thermoset composites functionally unrecyclable by conventional means.

Current recycling options are limited:

• Mechanical grinding: Produces low-value filler material with destroyed fiber properties
• Pyrolysis (500–700°C): Burns away the resin but degrades fiber properties by 30–50%, and is energy-intensive
• Conventional solvolysis: Uses harsh solvents at elevated temperatures and pressures

With approximately 30,000 tonnes of composite waste generated annually in Europe alone — and the first generation of wind turbine blades and aircraft structures now reaching end of life — the gap between "lightweight for sustainability" and "impossible to recycle" is becoming a reputational and regulatory crisis for the composites industry.

The Solution

Plastalyst is a catalytic hydrolysis process developed by AC Biode that decomposes polymers into their constituent monomers at temperatures at or below 200°C, using only water as a solvent [21][22]. No harsh chemicals. No extreme temperatures. No precious or rare-earth metal catalysts.

The process works on a wide spectrum of polymers — including challenging multilayer and deteriorated plastics that are considered unrecyclable by conventional means. AC Biode has demonstrated the ability to recycle PET back into terephthalic acid (TPA) and methanol — a true monomer-level recycling that enables production of virgin-quality material from waste [21].

The Numbers

Table 7: Composite recycling method comparison [21][22].

AC Biode is currently testing at 1.2-tonne scale for PET, PVC, PE, PP, and mixed plastics [21]. The extension to thermoset composite waste — epoxy/carbon fiber and polyester/glass fiber — is a natural next step given that the catalytic hydrolysis mechanism can break ester and amide bonds in the cross-linked network.

Why This Matters

If Plastalyst can deliver on its promise for thermoset composites at industrial scale, it resolves the composites industry's deepest sustainability contradiction. The same technology that makes structures lighter and more fuel-efficient would finally have a viable end-of-life pathway. Combined with P2M's GENE.SYS for material identification, you begin to see a complete circular economy infrastructure for composites: manufacture → track → identify → recycle at low temperature → reuse.

The Bigger Picture: Why the Startup Ecosystem Matters

Step back and look at these six companies together, and a pattern emerges:

Table 8: Six featured startups and the fundamental constraints they address.

Each startup attacks a different link in the composites value chain, but collectively they describe a future composites industry that is faster, greener, smarter, more flexible, and genuinely circular. This is not incremental improvement — it is structural transformation.

The Democratization Thread

There is a deeper story connecting these startups to each other — and to companies like Addcomposites.

The composites industry was built by and for aerospace primes. The materials were expensive. The equipment was expensive. The qualification was expensive. Everything about composites said: "This technology is for organizations with billion-dollar R&D budgets."

Every one of the six startups profiled here is pushing against that exclusivity:

AFP-XS on a KUKA industrial robot — aerospace-grade fiber placement, accessible to any lab or SME with a standard robot arm.

This is the same impulse that drives Addcomposites. When we built the AFP-XS — an automated fiber placement system that mounts on any industrial robot — the thesis was the same: aerospace-grade composite technology should not require aerospace-grade budgets. The technology should be accessible to university labs, SMEs, and Tier 2/3 suppliers that are building the future supply chain.

What to Watch on March 10

If you are at JEC World 2026, here is what we recommend:

References

Learn More