FLARE AEROSPACE

INSIGHT
FROM EARTH
TO US.

OBSERVE / UNDERSTAND / NAVIGATE / ACT

We develop visual localization and mission-autonomy systems for UAVs operating where GNSS cannot be relied on.

Explore technology
SCROLL TO EXPLORE

RESEARCH FOOTAGE / 3× PLAYBACK

THE WORLD
IS INFORMATION.

Read the world.

Overhead view of a school campus and its playing field.
Electronics and a cylindrical research prototype on a workbench.
Close-up of a quadrotor frame and its onboard electronics.
Drone and laptop prepared for a test on a playing field.
An overhead campus view with buildings and an athletics field.
FIELD ARCHIVE / 09Observation. Experiment. Flight.

A REFERENCE. EXTENDED.

Localization & Anchor System

HANSEL and GRETEL: research and reference-extension architectureIllustrative architecture: observations are compared with maps to evaluate position hypotheses. The intended reference-extension path connects to ground anchors from left to right. A shared dome illustrates proposed local coverage. HANSEL is a research prototype; GRETEL is under development, not field-validated.HANSELLOCALIZEENVIRONMENTSEMANTICSMAP ALIGNMENTGLOBAL POSITIONTRUSTED REFERENCEGRETELEXTENDGROUND ANCHORSLOCAL POSITIONING AREA

Illustrative: Reference found. Reference extended.

RESEARCH FOOTAGE / 3× PLAYBACKDrone footage over a school campus.Experimental archive. Not a product demonstration.

HANSEL

LOCALIZE

Research prototype

Localization research for map-referenced position hypotheses.

GRETEL

EXTEND

Under development

A reference-extension architecture using deployable UWB anchors.

INTENT BECOMES ACTION.

Agentic Autonomous UAV System

Mission input

Search this area and find black vehicles.

PROSPERO and ARIEL: a conceptual closed-loop missionHuman intent becomes area, target and task in sequence, then takeoff, transit, search, detect and report. The plan continues into flight. A vehicle observation returns to PROSPERO, replaces Search B with Search C, and ARIEL executes the revised route. Illustrative conceptual sequence, not live mission data or evidence of flight-controller integration.PROSPERODECIDEAREASEARCH REGIONSEARCH ATARGETBLACK VEHICLESDETECTTASKFIND / COUNTSEARCH BSEARCH CREPLANNEDTAKEOFFTRANSITREPORTDECOMPOSEPLAN / ALLOCATEMISSION PLANARIELACTABCBLACK VEHICLES / 01MISSION COMPLETE

PROSPERO / Receive

Receiving the search request.

Illustrative mission sequence. Not live telemetry.

RESEARCH ARCHIVE / PREPARATIONCanSat launch preparation. Separate research experiment.Experimental archive. Not a product demonstration.

PROSPERO

DECIDE

In development

Mission intelligence being developed to structure operator intent.

ARIEL

ACT

Execution architecture

An intended execution layer for flight, observation, and feedback.

TESTED. MEASURED.
WITH LIMITS.

Research evidence informs the stack. It is not a specification for every current product.

Predecessor research was evaluated across nine recorded real-flight trials at two physical sites. Reported horizontal localization errors ranged from 3.8 to 8.6 m, with a mean of approximately 5.5 m.

Limited predecessor field evaluation; not a universal performance specification or a current HANSEL performance guarantee.

Research approach

INSIGHT.
IN THE REAL WORLD.

Illustrative application concept: an operator and ground system overlooking a red-lit horizon.
Illustrative application conceptExplore Defense

FROM ROCKETS
TO AUTONOMY.

Flare's roots lie in hands-on rocket experimentation and rocket education.

A rocket-education business was registered in 2022. Early work included modular educational rocket kits and launches.

The work gradually moved from building vehicles to a harder question: how can an autonomous system understand where it is and what it should do next?

Today, Flare's R&D focuses on GNSS-independent localization and autonomous aerial systems.

THE WORK
BEHIND THE SYSTEM.

  1. Rocket experimentation
  2. Rocket education
  3. GNSS-challenged CanSat localization
  4. UAV semantic mapping
  5. Real-flight localization research
  6. HANSEL / GRETEL + PROSPERO / ARIEL

WHAT WE ARE
BUILDING NOW.

Flare is developing two connected technology families: map-referenced localization and reference extension, and mission intelligence and autonomous execution.

JEONGWOO KIM

Jeongwoo Kim leads Flare's technical development. His research spans UAV hardware, computer vision, semantic localization, and autonomous navigation.

  • 2025

    CanSat competition — team leader, top award in a two-person team.

  • 2026

    KSEF — Gold Award, individual research.

  • 2026

    Regeneron ISEF — Korea representative finalist.

Individual and team achievements, not Flare company awards.

R&D note: A Korean patent application related to GNSS-independent localization was filed in 2026. An application, not a granted patent.