Automatic Watering Systems in Vertical Farming Using the Adaline Algorithm

DOI

10.2991/aer.k.201221.070How to use a DOI?

Keywords

Automatic Watering, Adaline Algorithm, IoT, Vertical Farming

Abstract

This paper proposes a vertical farming model, by producing multi-layered plants that are stacked vertically. The research approach was carried out to obtain the technology used to achieve the goal of providing land at low cost, by utilizing the Internet of Things (IoT). The process of watering plants is transformed into an automation process with a sprinkler that adjusts to the calibration temperature, air humidity, and soil moisture value. The stages of implementation in this study will be directed to two processes, namely the data preprocessing stage and the Adaptive Neural Network training process. The Adaline algorithm will determine the duration of the automatic watering can be divided into two, namely the training process and the testing process. Process inputs and targets are trained with a network that has been built to add weight to learning then used based on incoming data training which is then used to facilitate the beginning or end of automation time and then this feature is used to determine the exact time the automation process is created effectively. Information about temperature, humidity, soil moisture, and when the sprinklers are activated can be monitored online via the internet with an application integrated with the IoT (Internet of Things) database. The application of Artificial Neural Networks (ANN), especially the Adaline algorithm, requires a knowledge base to be created using temperature, humidity, and soil parameters as parameters to determine the duration of automation. Watering duration is grouped into 3 types, namely short (5 seconds), long (10 seconds), and off (0 seconds). This knowledge base is also followed by the target value, plus input data that can be observed first which is then processed using normalization techniques, then the data with the Adaline concept can be implemented in an automatic watering system on vertical land. The test results obtained from the Adaline algorithm on an automatic watering tool obtained an accuracy value of 91.7% precision test results, then through the Mean Absolute Error Percentage (MAPE) validation test, an error value of 8.3% was obtained.

Copyright

© 2020, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Riki Ruli A. Siregar
AU  - Pritasari Palupiningsih
AU  - Inas Suha Lailah
AU  - Iriansyah BM Sangadji
AU  - Sigit Sukmajati
AU  - Novi Gusti Pahiyanti
PY  - 2020
DA  - 2020/12/22
TI  - Automatic Watering Systems in Vertical Farming Using the Adaline Algorithm
BT  - Proceedings of the International Seminar of Science and Applied Technology (ISSAT 2020)
PB  - Atlantis Press
SP  - 429
EP  - 435
SN  - 2352-5401
UR  - https://doi.org/10.2991/aer.k.201221.070
DO  - 10.2991/aer.k.201221.070
ID  - Siregar2020
ER  -