Helical Fusion targets 2027 power-on tests for Japan’s HARUKA pilot

The story

Japanese fusion startup Helical Fusion plans to begin preliminary power-on tests of its Helix HARUKA pilot device in 2027, marking an early checkpoint in an effort to turn decades of helical-confinement research into an integrated reactor platform. The company told Reuters that successful first-stage tests would support power-on testing of the completed facility around 2030. A commercial plant remains further away: Helical Fusion says its larger follow-on project could pursue net electricity in the 2030s and a power plant as early as the 2040s.

Those dates describe a sequence of engineering goals, not the arrival of commercial fusion. HARUKA is a compact demonstration machine intended to bring several essential subsystems together: high-temperature superconducting magnets, plasma heating, a blanket and divertor, and power-related equipment. The company’s own Helix Program separates that integration phase from Helix KANATA, a proposed larger device that would target 50 megawatts of net output. HARUKA therefore needs to prove that the architecture can work as a system before KANATA can test whether it can produce more usable power than the facility consumes.

The project uses a helical magnetic configuration, part of the broader stellarator family. Instead of relying on a large electrical current flowing through the plasma to help shape the confining field, twisted external coils create the required magnetic geometry. Japan’s National Institute for Fusion Science says this arrangement has attractive steady-state characteristics because the confinement field can be formed by external coils. The trade-off is formidable hardware: three-dimensional coils must be designed, manufactured and aligned accurately while operating under large electromagnetic forces and cryogenic conditions.

Helical Fusion’s design is rooted in the National Institute for Fusion Science’s Large Helical Device, or LHD, in Toki, Gifu Prefecture. LHD began plasma experiments in 1998 and completed its final experimental campaign in December 2025 after more than 200,000 discharges. It did not generate commercial electricity, but it provided a large research base in superconducting helical magnets, high-temperature plasma and long-duration operation. Helical Fusion was founded by researchers connected to that program and is building HARUKA at the same research site.

The immediate construction milestone is the first of HARUKA’s spiral coils. Reuters reported that work began in early 2026 and that the coil is now largely complete. Helical Fusion previously said a 2025 test of a high-temperature superconducting coil reproduced the magnetic environment expected inside a fusion device and sustained current under superconducting conditions. That component result was important, but an installed pair of coils must still operate with the cryogenic, structural, vacuum and control systems surrounding it.

The 2027 tests are expected to energize the early hardware rather than demonstrate a burning fusion plasma or send electricity to the grid. Later construction is due to add a second spiral coil and major components including a liquid-metal wall blanket. In a future power system, a blanket would have to absorb neutron energy, transfer heat and support the fuel cycle while surviving intense radiation. Those demands are among the reasons fusion timelines cannot be inferred from magnet performance alone.

The Japanese government is trying to build a broader industrial base around the field. Reuters reported that Japan plans 3.1 trillion yen in public and private fusion investment through fiscal 2040. That commitment can help suppliers develop superconductors, precision structures, power electronics and materials that multiple reactor concepts need. It does not remove the scientific and commercial tests confronting every developer: sustained plasma performance, tritium management, component lifetime, maintainability, regulatory approval and a cost of electricity that can compete with other low-carbon sources.

Helical Fusion chief executive Takaya Taguchi acknowledged that cost, safety and construction speed remain hurdles. That caution is important. After roughly seven decades of controlled-fusion research, no reactor has yet become a commercial power station. A machine can achieve a valuable plasma or magnet result without proving net electrical output, high availability or economical operation; each benchmark answers a different question.

INNOVOX analysis: HARUKA is consequential because it can expose the interfaces that optimistic road maps often hide. Magnets, blankets, heating systems and plasma controls may each pass separate tests yet fail to meet performance or maintenance targets when coupled. Publishing test conditions, failures and repeatability will be more informative than announcing a single successful energization. The project’s credibility should rise in proportion to independently reviewable operating data, not the ambition of its dates.

The next meaningful evidence will arrive in stages. First, the company must complete the coil set and show stable, repeatable electrical operation in 2027. Then it must integrate heating, vacuum and plasma-facing systems and explain how the liquid-metal blanket performs. Around 2030, a fully assembled HARUKA could show whether the helical approach is ready to scale. Only after that would KANATA’s 50-megawatt net-power target become a testable proposition rather than a design objective.

INNOVOX analysis

HARUKA matters because it moves a Japanese private-fusion program from component claims toward an integrated machine where interfaces can fail in ways isolated tests do not reveal. The decisive evidence will not be the first energized coil alone, but whether magnets, heating, plasma-facing components, heat removal and maintainability can operate together repeatedly. Until then, the project should be judged as a systems-engineering experiment, not a commercial power-plant breakthrough.

What to watch

Watch the 2027 coil and electrical tests, completion of the second helical coil, integration of the liquid-metal blanket and plasma-heating systems, and publication of independent plasma-performance and reliability data. By about 2030, the key question will be whether HARUKA can operate as an integrated device often enough to justify financing the much larger KANATA net-power demonstrator.