top of page


Autonomous platforms for enhanced mapping.

Autonomous platforms allow us to deploy innovative sensor technology in otherwise inaccessible regions. Unmanned aerial vehicles (aka drones) support high-resolution mapping of remote exploration areas, steep cliffs, or active open-pit mines. Terrestrial robots help explore confined or hazardous spaces such as underground mines and tunnels. 

As part of our mission to operationalize non-invasive material mapping, we combine the potential of autonomous platforms with innovative Earth observation technology. On the way, we have to address the challenges of implementing hyperspectral and geophysical sensors as payloads. These include 1) technical compatibility, 2) ensuring data quality under complex conditions, and 3) rapid and autonomous processing of large datasets. To achieve operational autonomy, we further research the potential of collaborative autonomous data collection by task-distributed drone swarms. It also motivates us to explore strategies for adapted sensor design, autonomous validation, real-time hypercloud processing, and improved autonomous navigation using hyperspectral information.



Development of innovative drone platforms. 


For maximum flexibility and interoperability with different payloads, we are designing, manufacturing and programming our drone platforms ourselves in our in-house DroneLab. This allows us to acquire drone based data with various payload categories, e.g. hyperspectral cameras in the visible, near- and shortwave infrared,  infrared thermal cameras, magnetometers, radiometers and Lidar sensors. Additionally we develop and test novel drone concepts, and share some of our drone systems as Open Hardware.


Processing workflows for innovative sensors


With the development of smaller and energy-efficient sensors, drones can serve as acquisition platforms for a variety of optical and geophysical sensors. However, this customizability also challenges the availability of reliable data processing workflows, including adapted tools for geometric and radiometric corrections. 


We want to contribute open, transferable and reliable tools for spectral and geophysical mapping using drone platforms: 

  • Processing of drone-based frame-based hyperspectral data → MEPHySTo 

  • Processing of  drone-based pushbroom hyperspectral data (nadir and oblique) → Hylite

  • Processing of drone-based magnetic and radiometric data → MAGNETO toolbox

  • Integration of surface (multi-/hyperspectral) and subsurface (geophysical, e.g. magnetic data) → Mulsedro Project, MOSMIN Project


Multiscale mapping

Drone data has the unique potential to bridge the scale gap between satellite or airborne remote sensing (with a pixel size of several to tens of meters), and the sparse but high detail of ground observations. We are currently exploring ways to better integrate drone and satellite data, to take advantage of the extent and relative availability of satellite data, while simultaneously retaining the high-fidelity spatial and spectral patterns captured in drone data.




Figure: Integrating multi-scale and varied extent datasets (a) using various learning approaches (b) to achieve wide-coverage super-resolution results (c) from satellite or airborne HSI (HyperUAV project).

This includes research aimed to:

  • Determine the influence of spatial scale on spectral derivatives (e.g. vegetation and geological indices, classifications) → HyperUAV Project

  • Compare remote sensing results over sensor types, platforms and scales to better understand the strengths and weaknesses of different platforms, and the sensitivity of different analysis approaches→ HyperUAV Project

  • Develop methods for upscaling and extrapolating data across scales, including e.g., super-resolution and generative resolution enhancement → HyperUAV Project

  • Demonstrate application to exploration for critical raw material mineralization, e.g. REE-bearing carbonatites 

  • Develop application for valorisation and monitoring of mine sites and mining related deposits → MOSMIN Project

Autonomous acquisition


Drone mapping often poses challenging trade-offs between flight time, data quality, overflown area, and legal constraints. We are currently developing an autonomous, task-distributed drone network to improve the characterization of isolated, remote targets in a time and resource efficient manner. We combine the advantages of light-weight systems (long flight duration, large overflown area) with those of heavy-duty systems (equipped with sophisticated sensors of high informational value). In-depth research on autonomous Multi-UAV systems, distributed coverage path planning, inter-drone communication and real-time data processing are at the core of our efforts (AutoTarget Project).


























Recently, we have been working intensively on the autonomization of the developed mapping routines for use in difficult-to-access or dangerous environments (e.g. mining shafts, tunnels). In addition to important preliminary investigations into the applicability of spectral imaging underground, our terrestrial robot platform REX was procured and equipped with a hyperspectral camera and a laser scanner. The hardware integration of the sensors and developments for data processing have been completed and the first underground data recordings have been successfully demonstrated.

Link to MSc thesis 






HORIZON-EUSPA Innovation Action - Multiscale Observation of Operations in Mines (2024–2027) 

Development of holistic, full-site services for the geotechnical and environmental monitoring as well as valorisation of mining-related deposits based on a combination of EO and in situ geophysical data. 

Partners: Helmholtz-Zentrum Dresden-Rossendorf e.V. (HZDR-HIF), Nordic Strategy Partners AB, Centre Tecnològic de Telecomunicacions de Catalunya (CTTC), GeoForschungsZentrum Potsdam (GFZ), Babes-Bolyai University (BBU), TheiaX GmbH, GeoKinesia, Boliden Mineral AB

Associated Partners: Advanced Mining Technology Center (AMTC), CODELCO, FQM Trident Ltd., CupruMin SA Abrud






CASUS Open Project - Scaling and spatial extrapolation of ultra-high resolution hyperspectral data (2022–2025)

Development of methods for integrating and upscaling high-resolution hyperspectral UAV surveys with/to other scale data sets, investigation of spectral mixing as variable of observation scale and development of generative deep learning methods for resolution enhancement of hyperspectral satellite data.

Partners: Helmholtz-Zentrum Dresden-Rossendorf e.V. (HZDR-HIF), Center for Advanced System Understanding (CASUS), Helmholtz-Zentrum für Umweltforschung (UFZ)






CASUS Open Project - Scaling hyperspectral data to meaningfully quantify essential ecosystem variables (2022–2025)

Utilization of the methods from HyperUAV-1 for a successful ecosystem management by developing data-driven models for essential ecosystem variables that bridge ground, UAV and satellite scale hyperspectral observations. 

Partners: Helmholtz-Zentrum für Umweltforschung (UFZ), Center for Advanced System Understanding (CASUS), Helmholtz-Zentrum Dresden-Rossendorf e.V. (HZDR-HIF)





CASUS Open Project - Autonomous multi-UAV (uncrewed aerial vehicle) for the characterization of remote and isolated targets (2022–2025)

Development of an autonomous, task-distributed drone network to improve the characterization of isolated, remote targets in a time and resource efficient manner. 

Partners: Helmholtz-Zentrum Dresden-Rossendorf e.V. (HZDR-HIF), Center for Advanced System Understanding (CASUS), Karlsruhe Institute of Technology (KIT-IIIT) 


Official Github Repository:





EIT RawMaterials upscaling project - Multi-sensor drones for geology mapping (2017 - 2020)

Development of two drone systems equipped with photogrammetry magnetic, and hyperspectral sensors for exploration, mining activity monitoring and post-mining rehabilitation.

Partners: Geologian tutkimuskeskus, GTK (Geological Survey of Finland), DMT GmbH & Co. KG, Geological Survey of Denmark and Greenland (GEUS), Helmholtz-Zentrum Dresden-Rossendorf e.V. (HZDR-HIF), LTU Business AB, Radai Oy


Publications and reports:

bottom of page