Madeiran agriculture is taking a new and decisive step in innovation, using new technologies to become more precise, efficient, of better quality and more profitable. In the specific case of bananas.
GESBA has developed a project called 'Precision Agriculture for Banana Cultivation - BAnana SEnsing (BASE)', which, among other objectives, aims to 'measure' the amount of water and nutrients the plant needs, as well as calculating the right time to cut the bunch. Technology and artificial intelligence are collabourating with humans in this traditional sector to improve resource efficiency and production management.
With the development of the project - which will enter a second phase once new EU funds have been approved, for which applications are already being prepared - it will be possible to predict the amount of fruit that will be produced in a given period, which will help GESBA to streamline human and material resources for managing banana stocks.
The aim of BASE is to "comprehensively analyse" the entire banana production cycle, using sensors placed in production and pilot agronomic studies to collect and validate information, which is then transmitted to a centralised data platform, subsequently analysed and used to monitor and optimise production and monitor its quality.
This goal of the Madeira Banana Sector Management Company has been achieved in partnership with Altice Labs, which has developed the technology for this purpose.
The newspaper got to know BASE at the Madeira Banana Center - BAM, with explanations from Diana Costa, GESBA's Quality Director and Filipe Freitas, a computer engineer from Altice Labs.
GESBA has other partners, such as the University of Madeira, which has studied the entire production cycle of the plant and understood all the growth dynamics, the data from which has been processed by a platform created by Altice Labs for this purpose. It also has ARDITI, which has developed technology based on Artificial Intelligence to know the right time to cut the banana bunch.
This is how the Artificial Intelligence/Machine Learning component of the project was exploited, with correlation algorithms between the parameters of the soil, plant and weather station sensors, said those responsible. "At the moment, the level of accuracy is 84%," said Filipe Freitas, commenting that the highest level achieved by specialists 'by eye' in the sector was 87%.
This project, which was started by GESBA and partners on June 1, 2021 and concluded on June 30 of this year, is being carried out at the Madeira Banana Center (BAM) and the Ponta do Sol Processing Center, where sensors, a weather station and 'Libelium' communication units have been set up next to the banana trees.
The first phase represented an investment of 552,541 euros, supported by Community funds under PRODERAM. Of this amount, 55,254 euros came from regional self-financing. The entities involved in the project will be the same for the second phase, which should also last three years.
Diana Corte and Filipe Freitas stressed that the impact of this project will be felt in the increased marketing of the banana product, reflected in an improvement in its quality and a reduction in its seasonality. It will also ensure that production, contrary to what is currently done, "in which two thirds of production is centered on the summer months, will also have an important weight throughout the year".
There is also the added value of attracting young farmers to this production, given its modernization and the control of various parameters through mobile phone, for example.
Filipe Freitas added that the station is made up of various elements, from the energy side with photovoltaic panels, a central station that receives all the information from the sensors that measure meteorological data such as humidity, temperature and light. There are also mpk' sensors in the banana tree's soil that allow us to know some of the banana tree's nutrients and thus interfere with the amount of fertilizer needed.
All this data is sent to Altice's platform, called 'LIVE! Green', which can be processed and monitored by GESBA. "We can infer, draw conclusions and apply some artificial intelligence algorithms to predict the growth and ripeness of the banana," added Filipe Freitas.
This information becomes even more important when we talk about the need to ration water consumption and climate change. "With the sensors, we can achieve these two things: in the case of water, when we should water and whether or not we are over-watering and wasting a water resource unnecessarily. In the case of the other sensors, which analyse nitrogen, phosphorus and potassium, for example, they allow for the rational use of fertilizers."
This means that the quality of the fruit is also improved, as well as the farmer's productivity and income. "The aim is also to improve yields on small plots like the ones we have in Madeira," said Diana Corte.
Attracting new farmers through technology
Among other missions, the BASE project wants to attract young people to agriculture. "Young people are much more connected to these new technologies and, by making the leap from the application created to a platform applied to a mobile phone, farmers will be able to monitor the data from the sensors and stations mounted on the crops.
According to Filipe Freitas, "the idea is to develop a kind of kit for the farmer, in which he himself, with his mobile phone and a set of sensors that he can install in his production, can understand the evolution of the banana's ripeness and also adjust the levels of water and fertilizer that he has to pour on his land".
By Paula Abreu
paulaabreu@jm-madeira.pt