Virtual sensors for smart farming: An IoT- and AI-enabled approach
Bibliographic record
Abstract
Smart farming relies on precise environmental data to optimize agricultural practices, with key metrics such as air temperature, humidity, rain, ambient light, ultraviolet (UV) radiation and soil moisture to play a crucial role in agricultural decision-making. However, the vast spatial coverage of agricultural fields and the high cost of deploying numerous physical sensors pose significant challenges, particularly for small and medium-sized farms. To address these issues, virtual sensors—machine learning models that predict sensor values based on data from relevant physical sensors—offer a cost-effective and scalable alternative. In this research, a number of Arduino-based IoT devices are designed and deployed equipped with various physical sensors, a lithium-polymer battery which recharges continuously using a 6 W waveshare solar panel, and a Real-Time Clock (RTC) module that synchronizes data logging. The IoT devices operated across two agricultural fields over a span of almost three months. The data collected form the basis for evaluating multiple machine learning models as virtual sensors. Furthermore, the use of open weather data to develop a hardware-free solution is explored. Experimental results show that virtual sensors provide a cost-effective and accurate method for replacing physical sensors. The Light Gradient Boosting Machine emerged as the most accurate model for virtual sensors, achieving prediction errors of less than 1% in most of the cases. This makes it a valuable tool for enabling cost-effective and data-driven farming in resource-constrained IoT devices. • Development of IoT-based virtual sensors for smart farming applications. • Cost-effective virtual sensors to replace physical sensors in agricultural fields. • Public dataset for training and evaluating machine learning-based virtual sensors. • Open-data integration for hardware-free virtual sensor modeling in agriculture.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".