Recap of what is up, for the engineering team of ROOTS
Our team, roots, has encountered many problems since this confinement. This post will relate them and explain the solutions that we found. Despite the fact that – clearly – we couldn’t do what we intended to (We needed to take some drastic decision to make the project sensibly easier) and that we became fairly late on the schedule, we believe that we still learned a lot from all this. We needed to intensify our innovative thinking to find ‘low-tech’ solutions to our problems. With absolutely zero electronic equipments, we successfully made first prototypes for our final sensor circuit.
We are also very sorry we haven’t wrote much lately. First we didn’t have much to write about (as we were waiting for deliveries, some of which didn’t arrive until now), and then we didn’t have much time to do it, as we had to catch up on the schedule.
Problem 1: absolutely no electrical equipment
We ordered some basic electrical tools – as a **soldering station** and some **basic components** – but it took several weeks for us the be delivered.
Then, we didn’t have any **voltmeter** or **current sensor**. Using an Arduino with some resistors, we could tune a very efficient voltmeter and current sensor to help us work with the project !
Problem 2: Use the wifi modules with a R-PI
We then ordered some wifi modules **ESP8266** to make a communication between the sensors and a R-PI that takes care of generating graphics depending on *when plants are touched* (to create this virtual graphic environment for museums, public spaces…)
When we received them, we quickly realized it wouldn’t be so easy.
– We didn’t have a voltage source good enough for them (not enough current goes out of the Arduino)
– We didn’t have any cable to ‘flash’ them, so it was impossible to either connect them to wifi or simply put some code on them.
We did a ‘home-made’ tunable voltage source from an old USB cable. Using a DC-DC converter that I found somewhere home. We successfully created a voltage source with enough current. So at this point, we could turn on the module.
Then, using a set-up that is as following
– Arduino connected to Computer with Serial communication
– ESP8266 connected to Arduino via UART
– Arduino sending all what it gets to the UART channel
we successfully created a way to send some firmware commands (AT Commands) to the ESP8266. It is very low-level, but it worked surprisingly well ! As the firmware was pre-installed, it was possible to communicate with it ! Without programming it, we could connect the module to the local wifi, and then to a server, and then stream data over it.
Then, we coded on the R-PI (that we already had since the beginning of this crisis) an UDP server (suitable for data streaming, with the *socket* python librairie) and we successfully connected the ESP8266 to the server. The good thing with this waiting time was that we could very well learn how to use a r-pi with SSH communication.
You can find some illustrations of what has been done here: https://github.com/arthurBricq/CHIC_work/blob/master/electronic_work.md
What’s next ?
This was obtained yesterday. We were planning to reach this state a lot earlier (at least 2 weeks ago) but deliveries are extremely slow and we encountered many problems in our path. Considering all this, we are pretty happy to have – at least – this working ‘fine’. As far as now, we have also learned many things even if we were not able to do everything as planned ! I strongly believe that I learned many things which I wouldn’t have if things would have went as planned.
Thanks to the *great workshop* done by Raffael with Octanis for PCB design and layout, we would like to create our first PCB by the end of the month. We are still lacking some elements (mostly the power supply and the sensing circuit) but we did some more orders and hopefully they will arrive faster than the previous ones. And we will keep attending the workshops of Octanis !
So the plan is to figure out everything for the power supply circuit, the sensing circuit, and then we would like to finalise a first version of our PCB.