GNSS-free drone development

Autonomous drone development for places without GPS

Revot Co., Ltd. (株式会社revot) · University of Tsukuba spin-off · Shinjuku, Tokyo

Revot Co., Ltd. is a University of Tsukuba spin-off in Tokyo. Revot develops drones that fly autonomously without GNSS satellite positioning and accepts commissioned development projects. Applications include equipment inspection inside factories, warehouses, and industrial plants. Revot handles localization, flight control, and LiDAR point cloud processing. An indoor flight proof of concept has an indicative starting price of JPY 980,000 and an indicative duration of 2–4 months, excluding consumption tax.

Inquiries in English are accepted by email at contact@revot.jp. The contact form on this site is in Japanese. Revot replies in English or Japanese.

Where satellite positioning is unavailable

Many drones determine their position using signals from global navigation satellite systems, or GNSS, which include GPS. Buildings and surrounding structures can block or disrupt these signals. GNSS-free autonomous flight uses other methods to estimate the drone's position in these environments.

  • Indoor spaces, including factories, warehouses, gymnasiums, and industrial plants.
  • Underground spaces, including tunnels, underground shopping areas, sewers, and mines.
  • Areas beneath bridges and elevated structures, and spaces surrounded by steel frameworks.
  • Areas between tall buildings, where reflected signals can distort position estimates.

Equipment inspection and infrastructure surveys often require access to spaces that people cannot easily enter. These spaces can also lack reliable satellite reception.

How drones estimate position without GPS

SLAM stands for simultaneous localization and mapping. It combines a map of the surroundings with an estimate of the drone's position. LiDAR measures distances to walls, columns, and the ground with laser light. Changes in the resulting point cloud help determine how the drone has moved. Cameras provide another approach to SLAM.

An inertial measurement unit, or IMU, measures acceleration and angular velocity. Combining these measurements with LiDAR or camera estimates helps maintain localization during brief sensor obstruction or sudden movement. This combination is called sensor fusion.

MethodSuitable conditions or roleLimitations
LiDAR SLAMOperation in dark environmentsSensor weight and cost; long, straight corridors with few distinctive features
Visual SLAMBright environments with visible patterns; lightweight sensorsDarkness, backlighting, and blank walls
IMUBridging short gaps in other sensor estimatesError accumulates over time; unsuitable as the sole positioning method
Barometric, ultrasonic, or time-of-flight sensorsMaintaining altitudeNo horizontal position information
Installed systems such as UWBIndoor locations where equipment can be installed in advanceSetup required at each site

Revot selects sensor combinations according to site conditions. Lighting, corridor length, and dust affect the design. The system must also determine which measurements to trust as conditions change.

Turning point clouds into usable data

GNSS-free flight can also collect detailed three-dimensional point clouds. Processing this data produces information for inspection and assessment.

  • Three-dimensional models of equipment and structures, including digital twins.
  • Comparisons with previous measurements to identify deformation, deterioration, or accumulated material.
  • Volume calculations for earthwork quantities or stockpiled materials.

The purpose extends beyond flight itself. Revot handles data acquisition and processing so that measurements from difficult-to-access spaces can support decisions.

Planning development around the site

  • Define site conditions: Lighting, available space, corridor length, dust, and radio conditions affect sensor requirements and development costs. A general-purpose drone may not operate as supplied at a site without GNSS.
  • Address flight and data processing: Stable flight requires control and localization technology. Turning recorded point clouds into usable outputs requires separate processing expertise. Development must address both.
  • Validate a limited area first: A proof of concept in part of the target site allows accuracy and practical operation to be assessed before expanding the scope.

For warehouse and factory projects, Revot first checks whether flight and image capture are feasible at the site. Revot establishes stable hovering and assistance for manual flight, then reviews the collected data before progressing to autonomous flight along a route.

Revot's development scope

Revot develops its own aircraft for autonomous flight without GPS. Rather than changing the settings of a commercially available aircraft, Revot implements the localization and control itself, so the design can address which sensor information the aircraft should trust at each site.

Revot accepts consultations covering site conditions, design, prototyping, flight validation, and operation. Technical work includes localization without GNSS, point cloud acquisition and processing, and three-dimensional reconstruction.

  • Indoor equipment inspection in factories, warehouses, and industrial plants.
  • Data acquisition for infrastructure surveys and disaster response.
  • Warehouse inventory drone development involving label or RFID reading.
  • SLAM and localization consulting for customers who carry out implementation themselves.

Three different GNSS problems

Loss of GPS can describe different conditions. Revot distinguishes these conditions because each requires a different response.

ConditionEffectDesign response
No reception indoors, in tunnels, or beneath bridgesSatellite positioning is unavailableUse localization that does not depend on external positioning
Intermittent reception between buildings, in valleys, or beside bridgesThe most dangerous case. The position estimate can jump when reception returns, causing sudden aircraft movementDefine behavior at signal loss and recovery, including transitions between positioning methods
Large errors in areas with signal reflectionsPosition estimates can deviate by several meters to more than ten metersCombine satellite positioning with other sensors to correct errors

In practice, intermittent reception is the most dangerous case. The aircraft may continue to trust an incorrect position estimate, and accidents occur here. Defining behavior at the moment reception is lost and the moment it returns is central to the design of GNSS-free flight.

Decisions to make before flight

  • Loss of localization: Define whether the aircraft should hover, land, or return along its previous route when it loses its position estimate.
  • Loss of communication: Define a separate response. Communication failure and loss of localization are different problems.
  • Flying over people: Decide whether the aircraft will fly over people. If it will, design the route first so that a falling aircraft would not hit anyone. Use propeller guards and set restricted flight areas.
  • Aircraft recovery: Plan how to retrieve an aircraft that stops in a location people cannot reach.

These decisions form part of the operating plan. Leaving them undefined can result in unpredictable aircraft behavior or an aircraft that cannot be recovered.

Development costs, durations, and inquiries

ScopeIndicative starting priceIndicative duration
Indoor flight proof of conceptJPY 980,0002–4 months
Localization implementationJPY 5,000,0004–8 months
Complete system including aircraftJPY 10,000,0006–12 months

All prices exclude consumption tax and are total project amounts before any subsidies. The figures are indicative; site conditions affect the required configuration and cost. Revot quotes SLAM and localization consulting separately for each request.

Revot accepts inquiries from the idea stage and offers a free consultation on flight feasibility in the proposed environment. Relevant documents or drawings can be sent to contact@revot.jp.

Revot accepts inquiries in English by email at contact@revot.jp. The contact form on the site is in Japanese. Revot replies in English or Japanese.

Frequently asked questions

What is a GNSS-free drone?

It is a drone that estimates its position and flies autonomously where satellite positioning is unavailable. Onboard LiDAR or cameras support SLAM, while an IMU provides acceleration and angular velocity measurements.

Can Revot develop a drone for indoor inspection?

Yes. Revot accepts development projects for factories, warehouses, and industrial plants. The scope can include aircraft design, prototyping, flight validation, and processing of the recorded point clouds.

Can Revot develop drones for warehouse inventory?

Yes. Revot accepts warehouse inventory drone development involving label or RFID reading. Revot first checks flight and image capture feasibility, then establishes hovering and manual flight assistance before progressing to autonomous routes.

How much does an indoor flight proof of concept cost?

An indoor flight proof of concept starts at JPY 980,000, excluding consumption tax, with an indicative duration of 2–4 months.

Can Revot provide only SLAM or localization consulting?

Yes. Revot can provide technical consulting while the customer handles implementation. Revot prepares a separate quotation for each request.

What can point cloud data show?

Point clouds can support three-dimensional models, comparisons of deformation or deterioration over time, assessments of accumulated material, and volume calculations.

When a drone flies where GPS is unavailable, what is the most dangerous case?

It is the most dangerous case, more so than a complete loss of reception. The aircraft may keep trusting an incorrect position estimate, and that estimate can jump when reception returns, causing sudden movement. Accidents occur here, so the design must define what the aircraft does at the moment of signal loss and the moment of recovery.

Can one sensor replace GPS?

No single method covers all the conditions described here. LiDAR, cameras, and inertial sensors have different limitations. The design combines methods and determines which measurements to trust under changing conditions.

Contact Revot

Inquiries in English are accepted by email at contact@revot.jp; the contact form is in Japanese. Seita Sukisaki, the representative director, replies directly within one to two business days.

Email contact@revot.jp

Revot Co., Ltd. / 3-16 Ichigaya-Tamachi, Shinjuku-ku, Tokyo 162-0843, Japan

Contact form (Japanese)