How the DOE fusion roadmap plans to close six critical gaps to the grid

Construction work at the CFS Devens facility, where SPARC’s magnet and vacuum vessel assemblies are progressing toward first plasma operations

(Image courtesy of Commonwealth Fusion Systems)

The U.S. Department of Energy has published its finalized Fusion Science and Technology Roadmap, setting out a national strategy for commercialising fusion energy by the mid-2030s. Built around a three-part Build-Innovate-Grow framework, the document identifies the critical science and technology gaps that must be closed before fusion pilot plants can reach the grid. More than 800 scientists and engineers from over 15 private companies, 10 national laboratories, and 72 universities shaped the strategy – but the document is explicit that execution depends on future Congressional appropriations, and it does not commit DOE to specific funding levels.

Six gaps, unequal urgency

The roadmap organises its work around six core Challenge Areas: structural materials, plasma-facing components and plasma-material interactions, confinement approaches, the fuel cycle, blankets, and fusion plant engineering and system integration. Each carries defined milestones and metrics, but the document’s own language signals that not all six are equal in terms of urgency or current readiness.

Structural materials and the tritium fuel cycle represent the longest-lead problems. The roadmap states that neutron-induced degradation is one of the most significant constraints on both reactor economics and safety, and identifies two critical knowledge gaps that remain open: whether existing material classes can survive fusion-relevant irradiation damage levels at all, and what their actual performance limits are. According to the document, the engineering materials database required to close those gaps is a prerequisite for designing and deploying any fusion power plant. DOE prioritises a Fusion Prototypical Neutron Source to begin closing this gap, placing full delivery on a longer-term timeline while pursuing near-term R&D to reduce the risk of that build.

On blankets, the roadmap is direct: no fusion blanket has yet been built and validated, and current testing environments for components are limited. Achieving a tritium breeding ratio above one – the threshold at which a fusion plant produces enough tritium to sustain its own fuel cycle – remains a prerequisite no fusion blanket has yet been built to meet. The document outlines a Tritium-Blanket technology Development Platform drawing on existing test facilities in the UK, Japan, Germany, and Canada as a near-term bridge, with a large-scale Integrated Blanket-Fuel Cycle Test Facility targeted in the longer term. For procurement specialists and supply chain professionals, the blanket and materials challenge areas represent the widest gap between current capability and what a pilot plant will require.

AI convergence and industrial scaling

The Innovate pillar integrates artificial intelligence as what the document terms a transformative enabling capability, specifically through an AI-Fusion Digital Convergence Platform. This will coordinate digital twins, surrogate models, and foundation models for plasma and materials science across the fusion ecosystem. Princeton Plasma Physics Laboratory is leading development of a fusion-centric supercomputing cluster called Stellar-AI, partnering with NVIDIA and IBM, to support foundation model training on fusion experimental and simulation data. The roadmap ties this directly to the Genesis Mission, a national AI initiative launched by executive order in November 2025, which specifically targets accelerating fusion energy delivery as one of 26 science and technology challenges identified in February 2026.

On the industrial side, the Grow pillar sets out public-private partnership mechanisms including the Milestone Program, which as of April 2026 includes eight companies working toward fusion pilot plant designs: Commonwealth Fusion Systems, Focused Energy, Realta Fusion, Thea Energy, Tokamak Energy, Type One Energy, Xcimer Energy, and Zap Energy. A new Fusion BRIDGE programme, beginning in fiscal year 2026, will co-fund experimental facilities with private partners, state governments, and philanthropies. The private sector has attracted more than $10 billion in cumulative investment, with $2.6 billion deployed in the twelve months to the end of 2025.

What the roadmap commits to and what it does not

DOE’s newly established Office of Fusion, created as part of a November 2025 departmental reorganisation, will lead implementation. According to the roadmap’s near-term timeline, first plasma from SPARC, the Commonwealth Fusion Systems high-field tokamak, is expected within two to three years. On the inertial confinement side, the National Ignition Facility has achieved eight successful ignition experiments to date, with the most recent recording an energy yield of 8.6 MJ from 2.08 MJ of energy input to the target.

Structured as a living document and updated at least annually, the roadmap is explicit that its milestones and timelines are contingent on future public-private partnerships and Congressional appropriations. For investors and procurement specialists tracking programme commitments, that distinction carries weight. The roadmap sets the agenda and defines the gaps. But the funding mechanisms that will actually close them remain subject to political and commercial conditions outside DOE’s direct control. The stated goal is to deliver the public infrastructure needed to support private fusion sector scale-up in the 2030s, and the roadmap’s value lies in defining precisely what that infrastructure needs to be.

Stay ahead in the fusion revolution explore more breakthroughs from leading innovators in clean energy technology.