KAUST scientists have used deep learning algorithms to accelerate the examination of thousands of years old, giant, stone rectangles in the Saudi desert.
“An international study showed that the huge, mysterious stone structures known as ‘Mustatil’ (Arab word for ‘Rectangle’) in northwestern Saudi Arabia, are among the oldest archeological ruins in the world,” Saudi Minister of Culture, Prince Badr bin Abdullah bin Farhan, said in a tweet in 2021.
These historic sites, which are around 7,000 years old, bewildered researchers and scientists who have long sought to determine their nature and the reasons behind their construction. A recent study by the University of Cambridge suggested that these huge structures, comprising chambers, entrances, and seats, are more complicated than expected.
‘Smart’ archeological survey
For quicker results, researchers at the King Abdullah University of Science and Technology (KAUST) have used an artificial intelligence network to carry out a detailed geological survey in the region, which hasn’t been sufficiently studied so far.
The team is composed of Dr. Silvio Giancola, researcher at KAUST’s Image and Video Understanding Lab (IVUL) and the Artificial Intelligence Initiative; Dr. Laurence Hapiot, archaeological research and cultural outreach fellow at KAUST; and Prof. Bernard Ghanem, IVUL senior researcher, and vice president of the Artificial Intelligence Initiative. The project is funded by the president bureau, dean bureau, and IVUL at KAUST.
AI tools are among the best methods used to assess archaeological sites and process general archaeological data, especially when it comes to spatial analyses such as the view field, which can be highly complicated without computers.
Rectangles of the desert
In 2020, the Saudi Heritage Commission announced that a scientific team discovered stone structures in the Nefud Desert, and identified the discovery as the oldest animal traps in the world, dating to 7,000 years.
According to the commission, the findings confirmed that the northern regions of the kingdom witnessed a cultural evolution in around 5,000 years BC. At the time, inhabitants built hundreds of large, stone constructions, which indicates cultural advancement in the region.
The fieldwork explored the archeological and environmental contexts of the stone constructions, especially the rectangle-shaped structure described as animal traps. These stone rectangles played a similar role and reflected a behavioral evolution that suggests a competition over pastures in complex, unstable environments in the Arabian Peninsula, even in periods of humidity like the Holocene era, during which people struggled with drought.
New research field
Inspired by a new research field known as ‘Computational archaeology’, this initiative used an AI software to model the exploration of stone structures with the help of satellites images.
Computational archaeology uses accurate, computer-based analytical methods including geographical information systems (GIS) to study data on long-term human behavior and behavioral evolution. Over more than a decade, archaeologists used available sources to manually analyze satellite images, and tools like Google Maps to search for possible archaeological sites.
In this project, KAUST’s researchers used automation to scan the unfamiliar, large rectangular stones in the Saudi Nefud Desert, in addition to other archaeological sites of circular and triangular shapes. The approach relies on machine learning algorithms fed with data sorted by Dr. Hapiot. Once the algorithms were trained, scientists became able to filter hundreds of similar characteristics on a wide scale. Now, when archaeologists discover a new structure, they can use the tool to convert similar pixels into geodetic data via GPS, and then combine results in a digital map and database for analysis.
“This demonstrates that KAUST is a unique research facility that excels in different faculties. Few environments can achieve an accelerated integration of deep, technical approaches like Artificial Intelligence in cooperation with archaeologists. This helped reach a different understanding of Nefud’s stone structures,” said Hapiot.
The extensively studied field in Nefud features thousands of massive, stone structures. Given that Saudi Arabia’s area is approximately two million square kilometers, geological surveys using conventional research operations and exploration methods could take months, or maybe years. But the new AI-based approach used by KAUST’s team took only five hours.
Commenting on the modern techniques used in this field, Dr. Jaser Suleiman al-Harbash, executive director of the Saudi Heritage Commission, said: “AI and machine learning processed huge sets of data from the Saudi archeological sites with an amazing speed. The commission hails the efforts made by KAUST to use the latest techniques in studying those ancient, stone structures. This can help us find more about the stones’ function and distribution, as well as the ancient civilization that built them.”
In addition to accelerating archaeological exploration, the new technique could provide answers to many questions about the size, capacity, and distribution of the stones, as well as determining whether exploring an ancient structure in a given region can help find other similar or linked structures in neighboring regions.
The benefits of the new deep learning technique used by KAUST are not limited to exploring archaeologic sites, as they can also help achieve the Vision 2030 goals, by preserving and documenting the unique heritage of Saudi Arabia, and promoting tourism. The new technique can be used in other regions with similar soil characteristics and topography. An initiative should be launched to help enhance the benefits of AI in archaeology, so archaeologists and data scientists can exchange their knowledge and achieve promising results.
Archaeology studies the whole activity of our ancestors in a given place and time. These activities include the tools made by humans to meet basic needs, construction, social and economic behaviors, written texts and architecture, and artistic and scientific works.
Archaeology also focuses on studying the origins of human civilizations, using the latest techniques that analyze the tiniest details related to our ancestors. The second half of the 20th century saw the emergence of the “New Archaeology” term, which indicates studying the organization of human communities in their locations, and defining their social structure in order to connect all these findings in a universal system on human behavior.