Towards Autonomous In-situ Soil Sampling and Mapping in Large-Scale Agricultural Environments
Thien Hoang Nguyen, Erik Muller, Michael Rubin, Xiaofei Wang, Fiorella Sibona, Salah Sukkarieh · 2025
Read Towards Autonomous In-situ Soil Sampling and Mapping in Large-Scale Agricultural Environments by Thien Hoang Nguyen, Erik Muller, Michael Rubin, Xiaofei Wang, Fiorella Sibona, Salah Sukkarieh (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-05653.txt.
Introduction
Checking soil across a big farm by hand is slow, costly, and too coarse for careful fertilizer planning. Thien Hoang Nguyen and co-authors present a robot system in this 2025 paper. The system samples soil in the field and tests it on the spot, building maps of soil health as it goes.
One unit pulls samples of set weight from a set depth, and a mobile lab unit measures pH and key nutrients in about ten minutes per sample. Trials on a large Australian farm showed steady sampling and fast, accurate tests. The system could give farmers quick, detailed soil data for smarter and greener field care.
Catalog record
| CREATOR | Thien Hoang Nguyen, Erik Muller, Michael Rubin, Xiaofei Wang, Fiorella Sibona, Salah Sukkarieh |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 35 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-05653
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Towards Autonomous In-situ Soil Sampling and Mapping in Large-Scale Agricultural Environments Thien Hoang Nguyen, Erik Muller, Michael Rubin, Xiaofei Wang, Fiorella Sibona, Salah Sukkarieh Abstract Traditional soil sampling and analysis methods are labor-intensive, time-consuming, and limited in spatial resolution, making them unsuitable for large-scale precision agriculture. To address these limitations, we present a robotic solution for real-time sampling, analysis and mapping of key soil properties. Our system consists of two main sub-systems: a Sample Acquisition System (SAS) for precise, automated in-field soil sampling; and a Sample Analysis Lab (Lab) for real-time soil property analysis. The system’s performance was validated through extensive field trials at a large-scale Australian farm. Experimental results show that the SAS can consistently acquire soil samples with a mass of 50g at a depth of 200mm, while the Lab can process each sample within 10 minutes to accurately measure pH and macronutrients. These results demonstrate the potential of the system to provide farmers with timely, data-driven insights for more efficient and sustainable soil management and fertilizer application. I INTRODUCTION Achieving sustainable agricultural resource management requires accurate, high-resolution, and up-to-date data on soil properties such as pH and macronutrients [ 1 , 2 ] . However, conventional soil sampling and testing methods fail to address this need at scale. A typical soil sampling and analysis procedure involves multiple manual steps, including sample collection, sending the samples to a laboratory facility, chemical treatment, filtration, and analysis. The sample collection step demands significant effort from trained personnel, who must perform sampling, store samples appropriately, and deliver them promptly to ensure accurate analysis results [ 3 ] . This process is not only time-consuming but also physically demanding, which becomes even more challenging as the number of samples increases. Furthermore, the results from laboratory facilities often take weeks to return, which further delays decision-making and makes these methods unsuitable for real-time, large-scale agricultural applications. Finally, conventional soil analysis methods yield discrete measurements of soil properties and nutrient levels at isolated sampling locations, which may limit farmers’ ability to optimize resource use and to enhance productivity. In contrast, dense, spatially correlated data can provide a comprehensive understanding of the complex relationships between critical factors such as pH, macronutrient levels, and crop health. A more automated and reliable solution, designed for large-scale operation, would allow such granular data to be obtained more effectively. Robotic solutions enable real-time, in-field assessments, reducing reliance on centralized laboratories and accelerating data collection for applications in agriculture, environmental monitoring, and land management [ 4 ] . Thus, there is a growing interest in developing robotic systems that can offer a transformative solution to the aforementioned challenges by automating sampling, preparation, and testing processes [ 5 , 6 ] . The crux of a robotic solution is its ability to measure the target soil properties automatically, reliably, and in situ. Unlike traditional laboratory-based soil testing methods, automation can reduce human labor, enhance efficiency, and improve the consistency and scalability of soil analysis. By providing rapidly acquired, high-frequency, and spatially dense data, robotic systems can support sustainable practices, minimize waste, and maximize productivity. (a) 3D model of the SAS (b) Real SAS sub-system (c) 3D model of the Lab (d) Real Lab sub-system Fig. 1 : 3D model and the real-world system in the field. In this paper, we present a novel robotic system capable of autonomous real-time sampling and mapping of key soil properties (Fig. 1 ). Our main contributions include: • Sample Acquisition System (SAS): A mobile robot equipped with an advanced sampling mechanism that can autonomously navigate to target locations, drill to precise depths, collect soil samples of specified mass, and return the samples to the base station; • Sample Analysis Lab (Lab): A mobile unit for on-site measurements of targeted soil properties, including pH and concentrations of Nitrogen (N), Phosphorus (P), and Potassium (K), featuring a semi-autonomous workflow to handle soil sample processing, data analysis, and environmental modeling tasks; • A modular system architecture comprising the SAS and Lab subsystems, where the SAS integrates a mobile robot and a drilling unit, both of which can be independently adapted or replaced to suit different terrains, soil types, and mission requirements; • Extensive real-life experiments on a 50,000 m 2 pasture farm validate the performance and robustness of our system under real-world farming conditions. II LITERATURE REVIEW Although various robotic solutions for in-field soil sampling or analysis have been put forth, most remain at the prototyping stage [ 6 , 5 ] . As such, we focus on systems that have been rigorously field-tested and validated, and examine the challenges that hinder their broader adoption. The automated on-the-go soil nitrate mapping system (SNMS) [ 7 ] and Agrobot Lala [ 8 ] are autonomous robotic systems designed to collect soil samples and analyze nitrate (NO {}_{3}^{-} ) levels directly in the field. Both systems utilize specialized actuators for soil collection and Ion-Selective Electrode (ISE) sensors for direct Nitrogen measurement. However, they differ in core functionality for the soil analysis task: while SNMS is entirely manual and lacks sample handling or processing capabilities, Agrobot incorporates both features. Nonetheless, Agrobot has several notable limitations that constrain its broader applicability. One significant drawback is its focus on analyzing only nitrate levels, limiting its utility for comprehensive soil nutrient mapping that includes other macronutrients such as Phosphorus and Potassium. Additionally, while the system provides real-time analysis, the process requires approximately 30 minutes per sample, which could hinder scalability for large fields requiring high-density sampling. The Soil2data project [ 9 ] is another innovative approach to automated, on-the-go soil testing for agricultural applications. Integrated with the Bonirob robot [ 10 ] , the Soil2data system is capable of autonomous navigation, sampling, analysis, and even fertilizer recommendation. The system employs a 16 mm auger, which can drill up to 30 cm to collect representative soil samples for nutrient analysis. The sensor module, equipped with ion-sensitive field-effect transistor (ISFET) sensors, provides real-time measurements of key soil parameters such as pH and NPK, with concentration levels up to 100 mMol/l. Soil processing and analysis are completed in approximately 15 minutes, with the ISFET sensors stabilizing within 100 seconds. Once testing is finished, the soil samples are left in the field, minimizing the need for transport and laboratory work. The NASA’s Curiosity rover, launched in 2012, features an advanced on-board laboratory called the Sample Analysis at Mars (SAM) [ 11 ] . This system is designed to analyze Martian soil and rock samples for the presence of key elements like carbon, oxygen, hydrogen, and nitrogen, which are essential for understanding Mars’ potential to support life. The SAM system is equipped with a sample manipulation system that transports powdered samples from the rover’s drill to the analysis chambers. This combination of tools allows the Curiosity rover to conduct comprehensive in-situ analyses of Martian soil. Even though the Soil2data and Curiosity systems represent significant advancements in automated soil analysis, several challenges remain in terms of reproducibility and widespread adoption. First, to employ the ISFET multi-sensor module, which is custom-made and not commercially available, Soil2data is meticulously engineered to ensure accuracy and reliability, addressing potential issues such as dust, light interference, moisture, mechanical stress, corrosion, and cross-sensor interference [ 12