Παράλληλα με το προσωπικό μου blog, διατηρώ με το συμφοιτητή και συνεργάτη μου Γιώργο Βαρβαρέλη και ακόμα ένα blog που είναι αποκλειστικά αφιερωμένο στο project της διπλωματικής μας εργασίας. Πριν λίγο, καθώς ήμουν καθ οδόν από Αθήνα για Βόλο, είχα πολύ χρόνο κι έκανα δύο σημαντικές δημοσιεύσεις σχετικά με το τι ακριβώς σκοπεύουμε να κάνουμε και στο δεύτερο ποστ μπήκα και σε πολύ πιο τεχνικές λεπτομέρειες σχετικά με τον αισθητήρα μέτρησης του δείκτη NDVI. Αναδημοσιεύω λοιπόν παρακάτω το πλήρες κείμενο καθώς και τον σύνδεσμο από το αντίστοιχο στο project blog. Θα μου άρεσε να λάβω κάποιο feedback για οτιδήποτε σκέφτεστε και θέλατε να συζητήσουμε.
As i am on the way to Volos from Athens, i think its the best time to explain how we are going to solve the first and fundamental problem of our project. How we are going to sense data from the field and what we are going to do with it in order to form some little and local fertilizer policies within the field.
Scientists have worked on this direction before us and they have developed a plant health index, called NDVI index. I will leave the explanation about how it works for later in the next post because i think the most important thing for somebody to understand is what it means. It shows us the health of the crops by measuring its amount of Green colour! After research, scientists have concluded in the opinion that the more green crops are, the less fertilizer they require as they have more than enough amount of Nitrogen. As a result, we have developed several commercial NDVI sensors who collect data from the field, create NDVI gis maps and they are used in fertilizer spreading operations as long as spraying operations. The problem is that NDVI itself is not enough to make us decide whether our crops have enough Nitrogen or not. Don't forget that Nitrogen is our one and only concern! NDVI index is just an index. It can easily makes us disoriented and take wrong decisions in some cases like having a sparse planted field! In this case, NDVI sensors measure the soil colour as long as wheat colour and the average of green and brown is something in between. The plants may be healthy with plenty of Nitrogen but if we consider NDVI as the only measurement, we will reach a false decision as we will apply more fertilizer on average than it is needed.
This is where our team will try to make an advance. Our project goal is to take more parameters into account and develop a more advanced "fertilizer amount decision system" based on both NDVI measurements and plant density and plant height. These 3 factors, we believe, are equally important when you need to make a decision how much fertilizer you should apply in a field. These 3 factors will be calculated in a real time data stream with the use of two sensors. The first one is Crop Circle developed by Holland Scientific and the second one is The SICK Laser Measurement Sensor (LMS) 200 which will help us define both plant height and density. All 3 data streams will be used as inputs to software we plan to develop, which will be responsible for dividing the field in zones. The number of zones we are going to use depends on the precision we want to have and on the data variation we will measure in practice. At last our goal is to develop a plug n' play piece of equipment which will be easily adaptable in every tractor even in old ones and collect data on the go and make small variations in the spreading machine for applying some more or less amount of fertilizer within a given range.
In the following Post, i will analyse how NDVI sensors work.... I would appreciate it if you send us some feedback and share your thoughts.