Automated Seeder Alert System

An alert system for a seeding machine used in semi-automated farms. This project is my task for my Internship at Cropx, an agro-based company.

This project is aimed to create an autonomous system that will alert the machine operator, if the seeds in the seeder machine are below a certain threshold. This will help vastly to overcome the problem where farmers usually have to manually check the seeder every few yards, increasing the seeding time and work. Moreover, if the operator misses to do so, they will have to unearth yards and yards of soil to check where the last seeds were dropped before the dropping stopped. With this proposed system, the farmer would receive an update even before the seeder is completely empty.

I have simulated a Ultrasonic sensor, HCS04, to alert when the parameter of 33cm, which is the depth of the seeder box.

In this edition, I have implemented the required circuit in hardware using Arduino UNO board, a HCS04 sensor and Breadboard. It can be seen as the object leaves the 40 cm boundary, marked by the pen, that the led lights up, indicating that nothing is present in between the Sensor and the object.

This mechanism will be mounted on top of the seeder box to monitor the seed levels in the box that decrease with the increase in seed flow. This affects the distance between the top layer of seeds and the sensor, hence the distance sensor can sense the seed levels.

Testing the alert mechanism on a seeder box and it can successfully detect the objects in between the sensor and the bottom of the seeder box.

Mounted the circuit onto an immobile stand for cross-directional analysis. The sensing has an intentional time delay of 2 seconds for mapping the depth of the seed box.

Added new features of a buzzer and a servo motor. The motor will be used to make an rotatable surface mount on top of the seed box so that the sensing mechanism can be "rotated away" from the top of the seed box.

Another version of the motorized sensing, now with a micro-servo motor for better control.

In this version, the sensing module was now attached to a PCB, Printed Circuit Board, for increased mobility and accuracy. The design is primitive and not aesthetic as this was made by the conventional, chemical etching process. The final version will be made by the new and better "buffing" mechanism through a PCB printer.

In this update, I have integrated a GSM module to send text messages from the machine to any valid mobile number. This will increase the autonomity of the seeder machine.

Due to difficulties faced with an out of box mechanism, a new approach with the use of lasers is tested. The results are promising.

 This time we test the laser prototype with a longer distance of more than 40cm to mark the accuracy of this system.

In this update, the laser is now bent with the use of a mirror so that the assembly of the LDR, i.e. Light Detecting Resistor, and the Laser on the same side of the equipment increasing adaptability of the module.

Here is the implementation of the new Laser based approach within the seeder box. As we can see the LDR sensor can easily differentiate between the object blocking and not blocking the Laser path.

In this update, we have now designed a 3D module to case the circuitry and set the laser diodes in position.

The use of both the 3D printed design and the Laser cut MDF sheet failed to provide the required support and accuracy to the laser diode placement. New and better design needs to be made.

A better prototype of the module casing seems to be working well and sets the laser diodes in position.

With the new and updated design, the laser diodes are now in place and what remains is to calibrate the position of LDR sensors on the opposite end.

With the LDR module in place, the Laser "cutting" is sensed at a good accuracy. Changes have to be made in the receiver side including inverting the digital logic of the signal so that we get an alert when the box is empty and not vice-versa. For this a simple inverter IC 7404 can be used. We can now proceed to real testing.

Finally, after weeks and weeks of testing, troubleshooting and extensive remodeling the #Laser based approached has worked to resolve the issue of Seeder box.

I have completed this project as a part of my internship at CropX. It was quite a fun and exciting journey but it had many dull and frustrating periods in btwn. Looking at the successful testing of the modules right now, it was all worth it.