HINOTATE Sets Out to Build a Domestic Drone Supply Chain

The incorporation date was August 21, 2026. HINOTATE is based in Tokyo’s Chiyoda Ward, has ¥100 million in capital and is led by Kohei Hayashi, a director of parent company Liberaware. Its stated business covers research, design, manufacturing, sales, operational support and maintenance for domestic dual-use unmanned aircraft; physical AI and related software; and a domestic supply chain extending to the components inside unmanned systems.

That final phrase is the heart of the announcement. A drone is visible as a finished object in the sky, but its availability is determined by layers hidden from view: flight controllers, motors, electronic speed controllers, propellers, cells, cameras, inertial sensors, positioning, radios, encryption, ground stations, cloud services and update software. Add production engineering, inspection, operator training, repair, spare parts and incident reporting. A supposedly domestic aircraft can still become unusable when one of those links breaks.

A platform larger than the aircraft

Calling a car domestic because its body was assembled locally would say little about its engine control, chips, batteries or service network. A sovereign drone capability cannot be measured only at final assembly either. Who controls the flight-control source code? Where does imagery travel, and who can push updates to the aircraft? If battery cells, magnets, chips or radio modules become unavailable, can engineers qualify an alternative? If a vulnerability is found, can the operator investigate it, correct it and deploy a patch across the fleet?

Liberaware’s June proposal placed that full system within scope. It said it would seek partners across components, communications and control, electronics, manufacturing, quality management and maintenance. Envisioned missions included disaster assessment, inspection of critical facilities, infrastructure monitoring, intelligence and surveillance, and systems that contribute to airspace safety. The board approved a dedicated subsidiary on August 3; incorporation was completed on August 21.

Layer	Capability required	Risk when it breaks
Airframe and propulsion	Structures, motors, controllers, propellers, batteries and thermal design	Production stops; endurance falls; substitute parts trigger new testing.
Sensing and links	Cameras, navigation, inertial sensors, radios, encryption and jamming resilience	Data leaks; links fail; the mission collapses without GPS.
Control and AI	Flight control, route planning, perception, ground stations and update management	Behavior cannot be audited, flaws cannot be fixed and services depend on outsiders.
Operations and sustainment	Training, procedures, maintenance, spares, incident response and data custody	Purchased aircraft sit idle or become unsafe in the field.

The real test of domestic capability is not “Did it fly?” It is “Can Japan field the needed number, repair one tomorrow and keep the fleet securely updated five years from now?”

Japan was not late to unmanned flight

Japan’s industrial drone story began over rice fields. As farmers aged and the burden of carrying pesticide tanks grew, an organization connected to the Agriculture Ministry commissioned Yamaha Motor in 1983 to develop an unmanned crop-dusting helicopter. After difficult experiments, Yamaha moved to a conventional main-rotor and tail-rotor design. The R-50 was completed in 1987, demonstrated spraying while carrying 15 kilograms and entered a closely monitored limited sale of 20 aircraft. Full marketing followed in 1989.

It was among the world’s first practical industrial unmanned helicopters. Large crewed crop dusters were a poor match for Japan’s small, scattered paddies. The R-50 grew from a specifically Japanese operating problem, then gained more sophisticated attitude control in the 1990s. Its successor, the RMAX, expanded beyond agriculture toward observation and survey work. Japan possessed early experience integrating machinery, control and real-world operations.

Yet the market’s center moved in the 2010s. A mass-production ecosystem led by Chinese manufacturers integrated compact motors, lithium batteries, cameras, smartphone-derived sensors and flight software at extraordinary speed. Prices fell, software improved and product cycles shortened. Japan had not forgotten how to build unmanned aircraft; it failed to translate its lead in large industrial helicopters into dominance of the new global multirotor ecosystem.

In 2015, rules entered the open sky

When modern small drones spread rapidly, Japan’s aviation law had no traffic code built around them. That changed quickly after a drone was discovered on the roof of the prime minister’s office in April 2015. Revised Civil Aeronautics Act provisions took effect on December 10, establishing prohibited areas around airports and densely inhabited districts, basic conditions such as daytime and visual-line-of-sight operations, and a process for obtaining permission for exceptions.

The regime later evolved from restriction toward managed deployment. In 2022, aircraft weighing 100 grams or more came within the unmanned-aircraft framework, while registration, airworthiness-related certification and pilot qualifications took shape. On December 5, Level 4 operations became possible under conditions, allowing beyond-visual-line-of-sight flight over populated areas.

This is where regulation meets the supply chain. Large public fleets need more than a prototype that passes a test once. A manufacturer must reproduce certified configurations, control component changes, record flights and maintenance, and support aircraft for years. Consumer-electronics culture—where parts can change quietly between production runs—does not fit public equipment that must remain traceable. HINOTATE is proposing to build the industrial discipline between those worlds.

IBIS2: a starting point inside pipes and rubble

Parent company Liberaware was founded in Chiba in 2016. Its specialty is not long endurance in an empty sky. It is flight inside places that are narrow, dark, hazardous and cut off from satellite navigation. IBIS2 measures 194 by 198.5 by 58 millimeters and weighs 243 grams with its battery. It captures inspection imagery in sewers, industrial facilities and ceiling spaces where sending a worker can be dangerous.

After the 2024 Noto Peninsula earthquake, IBIS2 entered collapsed homes and larger structures at risk of further failure, helping inspect damage and support utility restoration. In 2026, a separate 130-by-120-millimeter, 95-gram microdrone developed by Liberaware with Tokyo Electric Power support entered the primary containment vessel of Fukushima Daiichi’s Unit 3. It obtained images of what appeared to be the bottom of the reactor pressure vessel and nearby control-rod structures.

Those missions explain why HINOTATE sees continuity between civilian and security work. Rubble, pipes, reactors and critical facilities all combine weak communications, absent GPS, limited clearance and conditions too dangerous for a person. Miniaturization, collision tolerance, remote control, imaging and data analysis developed for disaster and infrastructure missions can also support security patrol and intelligence collection.

That is the meaning of dual use: a technical base can serve both civilian and security customers. The continuity also creates obligations. Who may buy the system? Which missions are accepted? How are exports and retransfers controlled? When autonomy is added, where does human judgment remain? Shared technology does not make all uses ethically or legally equivalent.

Why domestic supply became a national objective

In September 2020, Japan’s government agreed on procurement guidance requiring ministries, agencies and specified public bodies to consider cybersecurity risk when buying drones or contracting drone services. A vehicle photographing bridges, power plants, police facilities or emergency headquarters is more than a flying camera. It becomes a networked endpoint carrying locations, structures, operating patterns and potential vulnerabilities. A manufacturer or cloud operator may retain access through communications and software updates long after delivery.

From 2020 through 2021, NEDO funded standard aircraft and flight-control architecture intended for government procurement, along with key components and production support. Participants included ACSL, Yamaha Motor and NTT Docomo. The program produced SOTEN, a domestic aircraft aimed at government, local-authority, police, fire and utility markets. HINOTATE is therefore not Japan’s first attempt at a secure domestic drone. The next question is how to create a deeper supplier base and enough recurring demand to sustain it.

A Ministry of Economy, Trade and Industry interim report published at the end of 2025 estimated domestic unmanned-aircraft demand at roughly 140,000 units in 2030 and called for annual domestic production capacity on the order of 80,000 aircraft, including important components. In March 2026, Japan published a stable-supply policy for unmanned aircraft under its economic-security framework. HINOTATE was born into that policy tailwind.

“Domestic” is not a flag on the fuselage. It is the combined capacity to understand the design, trace the parts, substitute during disruption, patch the software and keep mission data under trusted control.

What physical AI means when the machine flies

HINOTATE’s stated remit explicitly includes “physical AI.” The term describes systems that do more than return an answer on a screen: sensors read the physical world, software chooses an action, machinery changes the environment and sensors measure the result. In a drone, that might mean estimating the distance to a wall in a dark pipe, holding position, identifying an anomaly and avoiding a collision even while the communications link is delayed.

A GPS-denied space is a severe proving ground. Indoors, cameras, inertial measurement units and ranging sensors must be fused to estimate position. Dust, steam, darkness, featureless walls and repeating pipe patterns can confuse perception. At a disaster site, even the map may no longer describe reality. A model trained in normal conditions is not guaranteed to perform on an unfamiliar ruin.

Public-purpose AI therefore needs a failure design, not merely an accuracy score. How much autonomy is permitted? What does the operator see? When the link is lost, does the aircraft stop, return or land? Can training data and update history be audited? Can the system resist jamming or deliberately misleading visual inputs? A trusted supply chain must include the ability to explain, halt and correct a “smart” aircraft’s decisions.

Dual use and the mass-production paradox

Domestic drones face a classic scale trap. Low volume keeps components expensive; high prices suppress demand. A business dependent only on defense can be pulled into low-volume, high-variation production as operational requirements change. One dependent only on civilian buyers must compete directly with inexpensive global mass-market aircraft.

The economic logic of dual use is to layer demand. If communications, flight control, power systems, image analysis and maintenance can be shared across power-line inspection, sewers, fire departments, police, coast surveillance and defense, research and factory overhead can be spread across a larger market. Daily civilian operations generate reliability data; exacting public requirements can improve quality for commercial customers.

Commonality has limits. A sewer drone and a maritime surveillance aircraft need different endurance, communications, waterproofing and payload. A cloud connection convenient for inspection work may be unacceptable for a sensitive security mission. The shared platform may therefore be less a universal aircraft than a set of interchangeable modules, validated software, parts-control rules, quality processes and training systems.

Why a former Rakuten CFO joined at the beginning

On the same day the company was incorporated, HINOTATE named Kenji Hirose—a former Rakuten Group representative director, deputy president and group CFO—to its board. Hirose began at Sanwa Bank, later joined Rakuten Securities and helped lead Rakuten’s financing and financial strategy, including during the buildout of its mobile business.

For a flight-technology company to add a large-company finance executive before unveiling an aircraft is revealing. Drone manufacturing absorbs money on both sides of the prototype. Components must be purchased, production and test equipment secured, certifications completed and long public procurement cycles survived. After delivery, the company must hold spares and trained technicians. Rapid order growth can increase working-capital stress rather than relieve it.

In a sensitive public business, governance also becomes part of the product: supplier screening, access to technical information, export controls, conflicts of interest, incident reporting and restrictions on use. Hirose’s immediate job is not to sell an unseen aircraft. It is to help build an institution customers can trust to keep supplying one.

Can HINOTATE define what “domestic” means?

Total self-sufficiency would be neither realistic nor efficient. A Japanese aircraft may contain overseas battery cells, semiconductors, sensors, magnets and open-source software. The goal should not be zero foreign content. It should be a precise map of dependencies: which part can halt a mission, under what legal and geopolitical conditions, and how quickly an alternative can be qualified.

Domestic procurement should not eliminate competition either. A protected market producing a few expensive aircraft will innovate slowly. Japan will need to benchmark international technology, develop jointly with trusted partners where useful and force domestic suppliers to compete on performance, price and delivery. Sovereign capability and national autarky are not synonyms.

The story begins after the first flight

Japan was one of the first countries to put industrial unmanned aircraft to work. The R-50 was not born from an abstract vision of an airborne revolution; it was built to keep people from carrying heavy chemical tanks through rice paddies. IBIS2 grew around another concrete task: do not send a worker into rubble or a dangerous pipe if a small machine can enter first. Japan’s distinctive strength lies in adapting technology to difficult, confined places where failure matters.

HINOTATE is trying to extend that field discipline into an industrial system. It has not shown its aircraft. Its suppliers, plant and orders remain undisclosed. Declaring victory now would be premature. Yet the sequence is significant: the company placed supply, maintenance, finance and governance in its mission before presenting a finished vehicle.

A drone makes news when it takes off. It becomes national capability after it lands—when someone changes its battery, protects its imagery, replaces a broken component, patches its software and keeps it ready until the next disaster. HINOTATE was established to build that unglamorous continuity.