DroneML

Year: 2024-2025

Supervision: Jitte Waagen

Execution and research: Jitte Waagen, Wouter Verschoof-van der Vaart, Agnes Schneider, eScience engineers: Christiaan Meijer, Maurice de Kleijn

Project description: 

The DroneML project aims to develop machine learning-based software that can rapidly screen multiple feature types (e.g., regularly shaped features that contrast with natural soil and grassland surroundings) and multiple multimodal input layers simultaneously, to enable rapid processing of large datasets for subsequent manual assessment of identified features. After initial identification, an ensemble learning approach such as ‘stacking’ may provide a machine learning supported metamodel for interpretation. The optimal solution would be if such a tool could be realized as, for example, a plugin in QGIS, a FOSS geographic information system, that can be run on a decent laptop, or could make use of cloud processing.

eScience Research Software Directory page

Publications:

Blogpost project announcement

DroneML project announcement

 

Jitte Waagen

Jitte Waagen has been granted a project in the 2023 Open eScience Call with DroneML, a project that seeks to improve the analysis of drone remote sensing data using machine learning approaches. In this project, the 4DRL team will closely collaborate with machine learning experts and software engineers from the eScience center, as well as two other researchers in the field (Wouter Verschoof-van der Vaart of the Nederlands Forensisch Instituut https://www.forensischinstituut.nl/ and Agnes Schneider, PhD at Universiteit Leiden https://www.universiteitleiden.nl/en/staffmembers/agnes-schneider#tab-1). This blog post is intended as a project announcement in which we will share the basic ideas of the project and what we aim to achieve.

Introduction

In the context of contemporary developments in climate change and human impact on the landscape, archaeological prospection techniques have become increasingly important. Due to decreasing groundwater levels, intensification of agricultural impact on the landscape and ongoing landscape development, important archaeological information is under continuous threat of being destroyed (e.g., Willemse 2021). Timely detection of buried archaeology is thus of essential value to its protection. Within archaeology, and heritage as a broader disciplinary context, remote sensing is part of an important set of non-invasive field techniques allowing archaeologists to search, evaluate and monitor archaeology in the landscape.

Drones in the Remote Sensing spectrum

In the field of archaeological prospection, drone/UAS (unmanned aerial systems)/RPA (remotely piloted aircraft) sensor techniques are very rapidly developing. Because of their operational flexibility they can very efficiently be deployed. In comparison with e.g., geophysical prospection techniques drones have a wider reach and can cover very irregular or inaccessible terrain. In comparison with remote sensing data derived from airplanes and satellites, their relative proximity to the ground is hugely beneficial because it allows for unprecedented high resolution and decreases distortions caused by atmospheric attenuation (e.g., Waagen et al, 2022). A technique such as thermal infrared imaging was of limited value for archaeological prospection until the arrival of drones. As such, the discipline has seen a swift rise in the deployment of drones using optical, thermal infrared, multispectral and LiDAR sensors, as an invaluable approach to detect, monitor, and evaluate archaeology.

Fast analysis required!

Due to the ongoing time-pressure, there is an urgent research challenge as to the analysis of the information captured by the drones. The deployment of state-of-the-art multisensor drone research strategies creates large multi-resolution and multimodal datasets of significant potential for archaeological sites and landscapes (e.g., enhanced visualisations of ground morphology, reflectance maps, vegetation indices). These layers add to existing satellite, aerial and terrestrial remote sensing data. The analysis of such datasets is a time-consuming process of inspecting every individual data layer one by one, marking potential anomalies, and interpreting them in a comparative analysis. The field of drone remote sensing data analysis would therefore advance fundamentally by the application of computer-aided inspection of these complex datasets.

DroneML, Machine Learning for multimodal and multilayered data

Therefore, DroneML aims to develop machine learning-based software that can rapidly screen multiple feature types (e.g., regularly shaped features that contrast with natural soil and grassland surroundings) and multiple multimodal input layers simultaneously, to enable rapid processing of large datasets for subsequent manual assessment of identified features. After initial identification, an ensemble learning approach such as ‘stacking’ (e.g., Wang et al, 2023) may provide a machine learning supported metamodel for interpretation. The optimal solution would be if such a tool could be realized as, for example, a plugin in QGIS, a FOSS geographic information system, that can be run on a decent laptop, or could make use of cloud processing.

The 4DRL has a lot of concrete use cases that can use as input and testing ground for the development of the tool; for example recent investigations of sites such as Weesp and Siegerswoude in The Netherlands (Waagen & van der Heiden 2021, Waagen 2023), but also abroad (e.g., Halos in Greece: LINK https://4dresearchlab.nl/drone-based-remote-sensing-at-halos-greece/).

Hopefully, the technology that would be the result of DroneML will hugely facilitate the work of archaeologists, as well as widen research possibilities, both within the field of heritage as well as any other discipline making use of remote sensing.

  • Waagen, J., & van der Heiden, M. (2021). Casestudy Siegerswoude-Middenwei. Thermisch infrarood remote sensing van een laatmiddeleeuwse nederzetting. In Archeologische prospectie vanuit de lucht.: Remote sensing in de Nederlandse archeologie (landbodems). (pp. 76-78). Rijksdienst voor het Cultureel Erfgoed.
  • Waagen, J., García Sánchez, J., van der Heiden, M., Kuiters, A., & Lulof, P. (2022). In the Heat of the Night: Comparative Assessment of Drone Thermography at the Archaeological Sites of Acquarossa, Italy, and Siegerswoude, The Netherlands. Drones, 6(7), [165].
  • Waagen, J. (2023). In search of a castle: multisensor UAS research at the Medieval site of ‘t Huijs ten Bosch, Weesp. 4DRL Report Series, 4. https://doi.org/10.21942/uva.23375486.v1
  • Wang, S., Wu, Y., Li, R., & Wang, X. (2023). Remote sensing-based retrieval of soil moisture content using stacking ensemble learning models. Land Degradation & Development, 34(3), 911– 925. https://doi.org/10.1002/ldr.4505
  • Willemse, N.W. (2021). Fragile monuments of the past, physical threats and countermeasures. (Nederlandse Archeologische Rapporten (NAR); No. 067). Rijksdienst voor het Cultureel Erfgoed.

 

4DRL project page

eScience Research Software Directory page

 

ARCfieldLAB: Where are we now?

Mason Scholte and Jitte Waagen

Introduction

In a previous blogpost we introduced ARCfieldLAB, a research project aiming at creating an inventory of the most important technological innovations of the last ten years in the field of archaeological remote sensing, and disseminating this knowledge to improve the quality of archaeological fieldwork and research in the Netherlands. In this blogpost, we provide an update on the achievements of the past few months, the current state of the ARCfieldLAB project and the progress over these past few months.

Achievements and Progress

Filling the knowledge base

The first of the two main components of the ARCfieldLAB project is the creation of an online knowledge base on innovative sensor technologies. In the course of the project, an extensive literature review has provided an overview of many ground-based, water-based, aerial, and satellite-based sensor technologies being used at the cutting-edge of international archaeological research. To highlight the value and potential of these sensor technologies for their application in the field of Dutch archaeology, as well as to provide an overview of resources and best-practices, several of these sensor technologies will be described in detail by ARCfieldLAB knowledge partners in the future knowledge base.

In the same vein, several case studies will take place to apply these innovative sensor technologies in Dutch contexts, both on land or in the water. The focus of these case studies is on ‘benchmark sites’: archaeological sites where the focus is explicitly on the application of various sensor technologies in combination. This will provide comparative opportunities that can provide validation of results, as well as a frame of reference to compare the strengths and limitations of each individual sensor. Further importance is placed on how these novel methods fit into the current AMZ-cyclus (the Dutch archaeological heritage management cycle).

For both the description of the innovative sensor technologies and the case studies, a call for projects was distributed to come into contact with potential knowledge partners. From the applications, a selection was made by the core consortium. The following knowledge partners were selected:

  • GAIA Prospection will be comparing the strengths of several different types of ground-based magnetometers at the site of a burial mound alignment in the province of Gelderland where various other geophysical techniques have taken place.
  • As part of a larger survey, Periplus Archeomare will be conducting a novel drone magnetometry survey on a sunken village in the province of Zeeland. Periplus Archeomare will also be responsible for the description of several maritime techniques: the sub-bottom profiler, side scan sonar, and multi-beam sonar.
  • Together, GAIA prospection and Periplus Archeomare will create the sensor technology description of magnetometry for both ground and water contexts.
  • ArcheoPro will return to one of their long-standing test locations for geophysical techniques in the south of Limburg at the site of a late medieval farm and conduct surveys using a wide array of sensors.
  • ArcheoPro is also performing a geophysical investigation in the center of a modern city to identify the medieval urban context underneath. This unique location provides a perfect opportunity to experiment with geophysical sensors in a highly complex sitecontext.

Several of the sensor descriptions and case studies are to be covered by knowledge partners from within the core consortium.

  • The 4D Research Lab will be continuing their multi-sensor drone-based investigation into the site of a 13th c. AD castle in Noord-Holland and a Late Medieval farmstead in Friesland  to further identify the environmental factors which influence the visibility of archaeological features. Given their expertise in applying thermography, the 4DRL will produce that sensor technology description.
  • Wouter Verschoof-van der Vaart will be advancing the application of the Dutch national LiDAR dataset in archaeological heritage protection through the use of change detection at various known sites in the Netherlands. The sensor technology description of LiDAR will similarly be produced by him.
  • Saxion will provide the technique descriptions for drone photogrammetry and multispectral research, as well as contribute to a yet-to-be-determined case study.
GAIA Prospection working on the magnetometry survey in the heath of the Veluwe. (Image credits: GAIA Prospection)
The 4DRL will continue with their drone-based analysis of a castle site in Noord-Holland (Image credits: Monumenten en Archeologie Amsterdam)
The first expertmeeting of sensor specialists within Dutch archaeology took place at the University of Amsterdam. (Image credits: 4DRL)

First expertmeeting

The second of the two main components of ARCfieldLAB is the organization of a series of expert meetings. The goal of these meetings is to promote the formation of multi-disciplinary networks and knowledge exchange collaboration between various sectors: archaeological professionals and academics, both from Dutch and international contexts, as well as remote sensing experts outside of the archaeological field.

On the 25th of April, the first of these expert meetings took place at the Bushuis at the University of Amsterdam. The goal of this meeting was to identify the needs and requirements for the online knowledge base, identify currently underexposed remote sensing techniques or sectors of untapped knowledge partners, and gain feedback on the direction of the project as a whole. All in all, the meeting was a great success and provided the opportunity for many interesting discussions. Various points were raised on the need for a decision support tool for sensor technologies, the importance of FAIR-data, and how sustainability of the project could be ensured after the project end date.

Looking forward

Second expertmeeting

On September 29th the second ARCfieldLAB expert meeting will take place at the Rijksdienst voor het Cultureel Erfgoed (RCE). The theme of this meeting will be ‘sustainable archaeological information structures’. The importance of this topic was highlighted in the first expert meeting. As ARCfieldLAB is involved in such a rapidly evolving field, it is important that the information provided is kept up-to-date throughout the project runtime. Furthermore, it is essential that there are plans in place for the maintenance and availability of the knowledge base after the project's duration ends. The focus of this meeting is therefore on how the information generated by the project can be made sustainable, and in what way this can be achieved through effective collaboration with the various E-RIHS projects. Stakeholders, experts on data management and sustainability, and members of related E-RIHS projects have been invited to discuss these issues.

Reuvensdagen

The sharing of knowledge is a vital part of the ARCfieldLAB project. For this reason, ARCfieldLAB will be showcased at the Reuvensdagen, the annual Dutch archaeological congress which brings together professionals, researchers, and other individuals interested in archaeology, within the context of a broader session on the various E-RIHS projects. The goal of this showcase is to increase the visibility of this project and obtain feedback to better integrate the final results with the current (and future) KNA (Knowledge Infrastructure in Dutch archaeology). The Reuvensdagen take place on the 16th and 17th of November in Hoorn. Interested? Find more information and tickets on the Reuvensdagen website!

ARCfieldLAB project announcement

Mason Scholte and Jitte Waagen

ARCfieldLAB. Innovative sensor technologies and methodologies for archaeological fieldwork: network, knowledgebase, and dissemination

BACKGROUND

The field of archaeological remote sensing has in the past decade seen significant developments in terms of novel sensor technologies and applications. These innovations can be applied to improve and expedite the archaeological fieldwork process in terms of the documentation, visualisation, and monitoring of archaeological features in a non-invasive manner, both on land as well as underwater.

With this blogpost, the 4D Research Lab presents ARCfieldLAB, a brand-new research project with the aim of creating an inventory of the most important technological innovations of the last ten years in the field of archaeological remote sensing, and disseminating this knowledge to improve the quality of archaeological research in the Netherlands. The project concerns a wide-ranging audience, including academic researchers, students, professional archaeologists and other specialists in this field (i.e. commercial companies or municipal and governmental archaeological services), and volunteers in archaeology.

This project is set to run for two years, and is funded by E-RIHS. E-RIHS is the European Research Infrastructure for Heritage Science which supports research on heritage interpretation, preservation, documentation and management. The mission of E-RIHS is to deliver integrated access to expertise, data and technologies through a permanent scientific infrastructure for heritage research, to which ARCfieldLAB will add a national digital platform for innovative methods and techniques and a collaborative network aimed at sharing experiences and best practices.

A core consortium led by the 4DRL of institutions firmly embedded in the Dutch archaeological sector or in the field of archaeological remote sensing has been appointed and acts as a steering committee this project. It consists of representatives of the Rijksdienst voor Cultureel Erfgoed (RCE), Stichting Infrastructuur Kwaliteitsborging Bodembeheer (SIKB), as well as the private sector (as represented by the Vereniging Ondernemers in Archeologie (VOiA)) and experts from Leiden University (LEI), the Free University of Amsterdam (VU), and University of Amsterdam (UvA).

An example of a recent innovation in archaeological remote sensing: drones are increasingly being utilized as remote sensing platforms.

THE PROJECT

There are two main components which constitute ARCfieldLAB:

The first component is the collection and dissemination of knowledge on innovative sensor technologies which can be applied to archaeological fieldwork by a) creating an overview of these developments in the last decade and sharing this knowledge through a publicly-accessible online knowledge base of resources and best practices, and b) providing examples of successful applications of the novel technologies and methods by which their value and potential for the archaeological fieldwork process is illustrated.

The second component is the organisation of a number of expert meetings, in which the possibilities and added value of innovative sensor technologies are elucidated and space is provided for experience in the application of these techniques to be exchanged. To promote multi-disciplinary collaboration, participants in these meetings will come from various sectors: archaeological professionals and academics, both from Dutch and international contexts, as well as remote sensing outside of the archaeological field. Additionally, workshops will be hosted for the promotion and education of these techniques.

As part of this project, various case studies will take place. These case studies serve to expose and fill in existing gaps in the knowledge of archaeological remote sensing in The Netherlands and aid in the development of best practices. The potential of the technological innovations which have so far not seen wide application in Dutch archaeology (but possibly have seen use in other countries or other sectors) as well as the efficacy of combining multiple remote sensing data sources in one site will be tested.

A graphical abstract of the ARCfieldLAB project.

EXAMPLE CASE STUDY: SIEGERSWOUDE

An example of a case study assessing the potential of a novel sensor technology in the context of Dutch archaeology is the use of drone-based thermal infrared remote sensing at the late medieval site of Siegerswoude, Friesland.

The theory behind thermography has previously been described in a previous blogpost, where it was used at the site of Halos in Greece. This pilot at Siegerswoude adds to a body of case studies which can be systematically compared to determine to what extent certain variables (e.g. soil composition, time of day, soil humidity, thermal properties of archaeological features) influence the outcome of an archaeological survey using drone-based thermography.

Historical sources associate the village of Siegerswoude, currently located on the meadow of a dairy farm, with a late-medieval grange from a regional Benedictine monastery situated approximately one kilometre west of the site. The site itself consists of at least five rectangular plots, evenly spaced along an axis and encircled by ditches.

Thermal imagery taken at the site revealed multiple traces which contrast with the background (marked A-E on the image) that have been identified as being archaeological in origin. The clearest of these is the rectangular ditch encircling the westernmost plot of land (A), visible on both the orthophoto as well as LiDAR data, which has a distinct thermal signature. On the northside of this feature, a double line is visible which is not present on the non-thermal data sources. Other traces included a rectangular trace in the centre of the western plot (B), long lines in SW-NE direction (C), and part of a ditch encircling the eastern plot (D) which continues into a similar double line feature as near (A). Test trenches have further validated these results, and provided insights into the use of this area: the ditches were used for draining the surrounding peat landscape, as well as for the extraction of loam.

One of the main takeaways from this survey, is the fact that thermography is capable of identifying archaeological features which are not visible on both orthophotos and LiDAR data of Siegerswoude. Furthermore, the noticeable differences in visibility of thermal signatures on the thermal imagery taken at different points throughout the day at Siegerswoude serves as a prime example of the importance of understanding the influence of variables on the results of these surveys.

The site of Siegerswoude as it is located in the Netherlands, together with optical imagery and AHN3 height data.
Thermal imagery from Siegerswoude showcasing various traces (indicated A-E).
References:
Waagen, J. & van der Heiden, M. (2021). Casestudy Siegerswoude-Middenwei. Thermisch infrarood remote sensing van een laatmiddeleeuwse nederzetting. In Archeologische prospectie vanuit de lucht.: Remote sensing in de Nederlandse archeologie (landbodems). Rijksdienst voor het Cultureel Erfgoed, p. 76-78.
Waagen, J., Sánchez, J.G., van der Heiden, M., Kuiters, A., & P. Lulof (2022). In the Heat of the Night: Comparative Assessment of Drone Thermography at the Archaeological Sites of Acquarossa, Italy, and Siegerswoude, The Netherlands. Drones, 6, 165. https://doi.org/10.3390/drones6070165
Rensink, E., Theunissen, L. & H. Feiken (eds.) (2022). Vanuit de lucht zie je meer. Remote sensing in de Nederlandse Archeologie, Amersfoort (Nederlandse Archeologische Rapporten 80).

3DWorkSpace project announcement

The NWO Open Science application that the 4D Research Lab submitted with main applicant dr. Jill Hilditch of the Tracing the Potters Wheel project (TPW), has received funding!

The project is an interdisciplinary collaboration on developing and deploying a 3D viewer for education and research purposes. This project, which we called 3D WorkSpace, is a collaboration with Loes Opgenhaffen, PhD researcher in the TPW project, Hugo Huurdeman, freelance designer (Timeless Future), Leon van Wissen, scientific programmer of CREATE, and is being executed in cooperation with Paul Melis and Casper van Leeuwen from SURF and the developers of the Smithsonian Institute represented by Jamie Cope, computer engineer at the Digitization Program Office of the Smithsonian Institute. Here, we present an outline of the project that will start in March, 2022.

Open access 3D models are often placed in online platforms with limited options for interactive communication and education. Although some 3D collections are published with their associated metadata, paradata, annotations and interpretations, these currently provide no open tools for re-use or interactivity. The Voyager digital museum curation tool suite, developed by the Smithsonian Institute, allows for interactivity and enrichment of the data but does not enable reuse or open content creation in a multiuser environment. Annotating, adding information to a 3D model without modifying the model itself, is possible for the creators of the content, but not for the end-users.

3DWorkSpace will facilitate re-use of 3D models through the addition of annotations and narratives, as well as side-by-side comparison of multiple models, within an online app environment adapted from the open source Voyager platform. It will allow data enrichment by enabling multi-authoring through the built-in annotation system, as well as through linkage to datasets (e.g., thesauri and museum catalogues) available as Linked Open Data (LOD). Two heritage-based case studies, production traces on experimental ceramics from the Tracing the Potter’s Wheel and a drone-based dataset from the 4D Research Lab, will allow full exploration of diverse 3D models for the implementation and testing of the adapted Voyager environment. Learning pathways, using the Voyager annotation feature, will train users in the necessary skills for guiding analysis of the 3D data models.

3DWorkSpace utilizes existing open access resources to realise a truly open science platform: from adaptation of the Voyager tool suite and testing with web-based open access 3D datasets, to technical support for data creation and access via Linked Open Data and Figshare. Evaluation will occur in tandem with the creation of training materials for technical set-up and 3D data curation. In this way we will lower the threshold for adoption, create best practice through training and demonstration, and create a roadmap for implementation and evaluation.

Although born from the material culture field, 3DWorkSpace is an initiative aimed at any field engaging with 3D data visualisation, as well as users seeking to integrate interactivity and data re-use, and will open up new ways of communicating and debating the narratives in which 3D reconstructions function for educators, researchers, students and general users.

We are really looking forward starting with this project!

Screenshot of the Voyager app