Case studies

Drone mapping with the eBee X discovers key archaeological sites in Italy

Duke University researchers recently used the eBee X with a multispectral camera in Italy to map a large archeological site.


Share | 01/27/2022

Not far from Rome, in the province of Orbetello, is Doganella (VII-III cent BCE), home to one of the most affluent and important Etruscan colonies of Etruria. Here, Duke University’s Professor Dr. Maurizio Forte leads a team of archaeologists who are implementing a new fully digital workflow to preserve entire excavation sites and their findings for greater accessibility worldwide.

The story behind how Doganella transformed lies in the buried structures and artifacts that await the team just below the surface. Using a combination of geospatial data recording, remote sensing and drone photogrammetry, Dr. Forte and Antonio Lo Piano, a doctorate student in classical studies, leads a research team in creating a 3D replication of the entire site – and there’s a lot of ground to cover.

Project Details 
Mission Type: Archaeology
Location: Italy, Vulci (Montalto di Castro) and Doganella (Orbetello)
Area: 700 ha / 1,730 A
Drone: eBee X
GSD: 4.8 in/pixel @ 60% overlap
Flights: 20
Total Images: 15,000
Processing Time: Weeks
Processing: Pix4D, Metashape
Outputs: Orthomosaic, DTM, NDVI, multispectral maps, 3D models, point clouds
Data by: Duke University

New field, faster results

Doganella, combined with its sister site Vulci, near the town of Montalto di Castro, is roughly 700 hectares / 1,730 acres and features terrain which resembles a modern-day farm field. Both archaeological sites have not yet been mapped or fully investigated and it’s estimated that 95% of the archaeological remains are still buried and only accessible through remote sensing.

The archaeological site of Doganella in the province of Orbetello, Italy has all the appearances of a modern-day farm field vs the flourishing Etruscan colony it once was. Photo: Duke University

Dr. Forte, a longtime advocate for using digital methods for archeological fieldwork, knew that a drone would help the team to collect data and identify areas of interest while guiding efforts for protecting the site and future excavations.

The team’s main objective from June through December 2021 was the creation of classified maps of the landscape with accurate identification of soil marks and crop marks. Soil marks and crop marks usually identify the outline of archaeological buildings, graves, roads, or other infrastructures.

Duke University archaeology team, (L-R) Nevio Danelon, Photogrammetry Specialist, Dr. Maurizio Forte, and Antonio LoPiano, Ph.D. Candidate in the field with the eBee X.
Photo: Duke University

Aerial drones and especially long-range fixed-wing models like the eBee X can rapidly cover the entire surface area of a large-scale archaeological site. By comparison, the amount of coverage required on the Doganella project alone would take weeks or months of traditional field walking surveys.


“The use of multispectral drones in Etruscan archaeology is really rare and these results are outstanding. In a relatively short time, we generated thousands of images and data with very high accuracy and with several outputs: digital terrain models, point clouds, orthomosaics, spectral mapping, 3D models. In short, the creation of a multilayered archaeological landscape never seen before.”

Dr. Maurizio Forte, Duke University


Multispectral mapping potential

Another challenge is that the archaeological landscape in the summer is overwhelmed by vegetation and crops making it difficult for the team to recognize sites and archaeological features in RGB imagery alone.

The use of a 5-band multispectral sensor can classify the landscape by infrared and NDVI and identify archaeological features related to the percentage of humidity, vegetation and crop growth.

Dr. Forte has flown drones for years and to meet the demands of this expedition he partnered with AgEagle to use an eBee X paired with a multispectral sensor.

Over a few weeks, the team flew the eBee X 20 times, collecting 15,000 images with 60-70% overlap at an accuracy of 4.8 in/pixel – 12 cm/pixel. The images were then processed in Pix4D and Metashape to create a variety of outputs.

This digital surface model shows elevation variances across the terrain.
Photo: Duke University

“The use of multispectral drones in Etruscan archaeology is really rare and these results are outstanding,” says Dr. Forte. “In a relatively short time, we generated thousands of images and data with very high accuracy and with several outputs: digital terrain models, point clouds, orthomosaics, spectral mapping, 3D models. In short, the creation of a multilayered archaeological landscape never seen before.”

In this hillshade of the main natural terrace, city walls can be seen in the south and northeast indicating the largest section of the circuit wall and the city’s major gate. Photo: Duke University

Common in early settlements, defensive walls visible in the terrain morphology (N and E side of the imagery), served to protect the colony against outsiders. These perimeters cannot be seen in RGB imagery, however, are easier to identify thanks to the combination of multispectral imagery and digital elevation models.

Looking at a section of the city wall in the northeast, the dark purple areas indicate mounds, and the area in between is the robbed-out city wall. Photo: Duke University

“The multispectral sensor is cutting edge and the inclusion of 5 bands provided outstanding results,” says Dr. Forte. “Our team was able to identify numerous individual tombs, structures, and large sections of the ancient street network still buried in the fields of these sites thanks to the capabilities of the camera.”

Several blocks of rectangular structures in purple can be seen here in between the streets marked out in yellow. The darkest regions of the NDVI are also related to the densest area of artifact surface scatter, including tile and pottery fragments. Photo: Duke University

This small region of the settlement shows a grid of buildings all in the same orientation or possibly a larger complex. In time, future excavation will reveal their purpose.

Two sensors are better than one

Dr. Forte discovered that a combination of sensors has proven particularly successful in several sites of interest. The imagery collected by the eBee X will serve as a precise guide for a secondary team using ground-penetrating radar (GPR). Multi-modal remote sensing will extend the team’s perspective of the site by obtaining a secondary reflectance map.

However, to make a GPR map of a whole 300 to 400 ha / 740 to 988 ac area might take months; with the eBee X, the team collected a site-wide map in one day.

“The eBee-X is the best choice for large-scale projects and the multispectral sensor due to its battery life, software and hardware integration, and portability of the system,” adds Dr. Forte. “Because of the fixed wings and the multispectral camera, it is ideal for mapping and classifying large archaeological sites.”

The archeological site of Doganella, seen here in the eMotion flight planning software, is 8 m / 26 ft above sea level. The team mapped this area with the eBee X at 70 m / 230 ft ATO.
Photo: Duke University

“The integration of the hardware with the eMotion software greatly simplified flight planning,” says Dr. Forte. “The 3D maps and integrated weather and wind data enabled us to make more informed decisions both before and during the flights. It is easily the best flight planning software we have worked with.”

The use of a drone before bringing in other equipment has both time and cost-saving advantages. GPR equipment is costly to transport and operate and having a general multispectral map of the entire site from drone enables the team to bring in the GPR on a very small scale and image specific areas of interest with great precision.

Environmental & cultural clues

Another interesting feature discovered within the multispectral imagery is that the orientation of the Etruscan buildings is very different from the orientation of the agricultural fields, which means the organization of the territory was different from the medieval, late-Roman period. It’s very likely the Romans started these changes, and the topography evolved into the Tuscany landscape we see today.

There are very few features such as dark soil, which you can see in the RBG image to the right, but when compared with multispectral imagery (left), and a blue band (left bottom), features such as the building and road marked by the red arrow, become visible. Photo: Duke University

Thanks to the results, Dr. Forte’s team can plan future archaeological campaigns for excavations in specific areas identified by the eBee X drone. This work continues to completely redefine these sites and in the case of Doganella, its long urban transformation.

This project is made under the permit and in collaboration with the Soprintendenza ABAP Siena Grosseto e Arezzo.

Newsletter sign up

Sign up to receive updates directly to your inbox.