Hello my name is Peter Ergisi, I currently study Mathematics, Further Mathematics, Physics and Computing at A Level. Nuffield Research Placements provided me with an opportunity to work along with the professional here at n-Coders, within four weeks of arrival I went from having no experience with STEM projects to creating a commercial product. Here is my story.
The first week at n-Coders was my hardest; I had never been in a professional business environment before. Nonetheless, I felt at home almost immediately; the staff had an extremely warm welcome they treated me like a part of their growing family.
Jamie (Quality Assurance Manager) mentored me throughout my stay. Although, the other members of staff were always willing to improve my skill set and knowledge. Jamie made me realise aspects of STEM projects that I would have never thought of before.
The purpose of my project was to create an Automatic Number Plate Barrier system for the Innovation Center that I was working at with n-Coders. Initially, my first instinct was to jump straight in and start building. However, Jamie made it very clear that my approach just wouldn’t work. I would fail! Clearly he knew something that I didn’t. He went ahead and demonstrated how to create a specification for my project, it soon became apparent why. I found solutions to problems that I hadn’t even thought existed. Eventually, by the end of the week, the specification allowed me to create a complete plan for the upcoming week.
Initially, I started exploring with an Arduino YUN board, a device that I have never used before. The first challenge I faced was providing power to the board, it seems the MicroUSB connector was damaged, which required me to connect a USB cable straight to the solder points on the board. Luckily, I was up and running within minutes of setting up the soldering station.
Here you can see that in two weeks I had an Ultrasonic Sensor working with an Arduino, all programmed in a whole new programming language to me (C). The sensor would illuminate an alternate LED when it detected an object is moving away from it, and vice versa.
However, I soon hit a roadblock; the Arduino YUN was great for ‘C’ based applications, but the limited Linux environment was an issue for me. The problematic Python instance determined that I wouldn’t be able to run the necessary application for the project, such as a Dropbox uploader script. To conclude the week, I had picked up some Arduino/C experience but ultimately had to switch over to the Raspberry Pi where I was more comfortable.
Additionally, along with my project work I was given the opportunity to execute real office work; I carried out some website testing, deciphered how to use a very confusing POS system and researched Pivotal Tracker solutions (Job tracking).
This week was exciting, to a great extent. The hardware we ordered had arrived for me to tinker with (USB Webcam). My initial task was to get the Ultrasonic sensor working on the Raspberry Pi, which was more than easy enough. Next, I had the USB Webcam working with the Raspberry Pi. My goal was to take a picture when the Ultrasonic sensor reported a certain distance. After a couple of hours I had fabricated a Python script to: probe the Ultrasonic sensor, detect the distance to the nearest object, compare it to a set value (50cm), take an image if necessary, upload to Dropbox and delete the image.
Succeeding the hardware, it was best I got started designing an enclosure.
Week 4 Building
After designing and picking up the materials, I started the building process.
I took apart the USB Camera and removed the enclosure because it was unnecessary and bulky. I then connected the Raspberry Pi to USB power directly because the power input was on the side, which caused issues with the space available in my enclosure.
Next we cut the square tubing to the appropriate length and mounted it on a gutter stopper. The gutter stopper was a temporary solution to keeping the system up. On commercial applications, it would have a different base.
Next I measured the height of the number plate for all the cars in the car park and created an average. This allowed me to drill holes into the tubing and place the camera at the perfect capturing height for number plates.
However, I soon encountered a problem. I was unable to place my electronics into the tubing because the holes were extremely inaccessible from either end of the tube. I decided to create an access door on the back of the tube.
While creating the access door I also created two cut outs at the top of the tube, this would allow me to place LED’s and give the car user an indication of processing etc.
To finish the enclosure and build, I glued all the components into the enclosure and create an all in one system.
I then connected the system to the internet and started testing the device. When a person approached the sensors, the system took a picture, processed the image and gave an output for the number plate. All while also uploading to Dropbox, updating the web panel and providing a functioning database for cars.