Physik | Technik
Karim Malik, 2004 | Zug, ZG
Frederick Steinmann, 2005 | Flims Dorf , GR
With the increasing risk of food insecurity worldwide, aquaponic gardening emerges as a productive and reliable means of ensuring fresh produce is available year-round. In this paper, a Western African community where water scarcity is a principal concern and fishery is no longer sustainable, is used as an example for the implementation of aquaponics. However, due to the high cost of adequate control systems for this application, the financial barrier of entry is unrealistic for such developing communities. In this research work it was hypothesised that a low-cost and modular solution could be implemented using a single-board-computing approach with common micro-controllers. Past literature is examined, shortcomings are noted, and a finalised solution utilising a node-based architecture built on ESP-32 and Raspberry Pi platforms with cloud based databases is presented. A novel, fully modular, and uniquely affordable IoT-based approach to tackle the high costs of aquaponics control units is demonstrated for real-world use. The design is operated with well-tested open-source frameworks, namely Node-RED and the industry-standard MQTT IoT protocol for communication and management of actuators and sensor readings. Further expandability and additional functionality is considered along with practical implementation challenges.
Introduction
What specific hardware and software standards are necessary to enable seamless integration of a low-cost and modular aquaponics sensing and actuating unit with existing monitoring systems and databases?
Methods
The setup included a Raspberry Pi 3B hosting a WAP to which the ESP-32 microcontroller connected to over WiFi. To test the MQTT protocol data broadcasting capabilities, a BME280 sensor was connected to the ESP-32 using I2C. The PubSubClient C++ library was used with the ArduinoJSON library to accommodate JSON parsing. The low-level computing unit or node was designed with Altium CircuitMaker, simulated in CircuitLab and incorporates a recent ESP-32-S-WROOM main computing unit, as well as a Texas Instruments LMP-91200 AFE for signal processing. Cloud communication is achieved using a LTE modem.
Results
The Raspberry Pi 3B supported the software with memory utilisation peaking at 250 MB. Low-level sensor interfaces worked reliably, and JSON-formatted data from the BME280 sensor was stored in the InfluxDB database for future time-series analysis. Alerting features were implemented with Node-RED and Grafana. The scripts for sensor and mesh integration worked as expected, with an included calibration function. Peak power draw of one low-level node did not exceed 300 mA while under peak load, and as such power efficiency is satisfactory. Computing power was recorded to have reasonable compute resources even with all sensors attached.
Discussion
The Raspberry Pi Zero 2W did not meet expectations despite its reduced cost, as the 250 MB memory consumption would not allow for software expandability. Latency was present when connecting to the web clients and accessing the database, indicating that the minimum specification would require 1 GB of memory or higher, leaving the Pi 3 or CM4 as suitable options. The cost of one node is now only 30 CHf, being 90% lower than comparable options. However, sensor electrodes remain a concern for cost-effectiveness, costing over 300 CHf, making up a majority of a unit’s cost. To improve ease of use and affordability, the cost of such sensors must be reduced appreciably.
Conclusions
The results discussed indicate the feasibility of implementing this system using commonly available components. A unique aspect of this design is its modularity, enabling effective adaptability to different climates such as that of Ghana. The potential social impacts of this design primarily include reduced water consumption for agriculture, leading to a higher proportion of clean water available for sanitation. Further, increased economic opportunities in the community follow from the scalability and profitability of aquaponics systems, potentially providing dozens of jobs at scale. Societally, it would contribute to lower costs and better overall nutrition in the community by providing year-round fresh produce. The barrier of entry for implementing aquaponics on a larger scale in West Africa or other less economically developed areas is made less complicated because of the platform’s uniquely high level of scalability and affordability. Moving forward, lower-cost electrodes must be pursued to ensure greater feasibility for economically disadvantaged communities, which may be possible through manufacturer collaborations.
Würdigung durch die Expertin
Elisa Bossi
This work addresses the important issue of food security by investigating easily applicable and large-scale scientific Aquaponic gardening Control Unit systems in the West African setting. In the analysis of the results, fundamental aspects for the system implementation in the local setting such as management complexity or costs were also considered. The authors showed great motivation in the project knowing that their results could be used in practice and be the starting point for further research in the field of aquaponic gardening in developing countries.
Prädikat:
sehr gut
Sonderpreis «Stockholm Junior Water Prize» gestiftet von der Swiss Toilet Organisation & Swiss Water Partnership
International School of Zug and Luzern, Hünenberg
Lehrer: Christopher Koch
