Using a Raspberry Pi instead of Beagle Bone


Are all the sensors mentioned here even compatible with Raspberry Pi?


Your question is very vague. Which sensors are you talking about? If you want a meaningful answer, it would help to explain more about what you’re trying to accomplish.


I have thoroughly gone through your website. I see that the codes which have been uploaded on GitHub are for BeagleBone Black basically. But usage of BeagleBone isn’t really cost effective for us and we are using RaspBerry Pi and it’s associated compatible sensors. So what my question is "Is there a Raspberry Pi version of the code or else can you please guide me through modifying the existing code to make it compatible with the RaspBerry Pi.


Based on your wording, it sounds like you think you’re making a reasonable request. That’s understandable. But, what you’re asking for would actually be a huge amount of work that would include PCB layout modifications, lots of discussions about your requirements, and plenty of software work–the sort of thing you could expect a paid consulting firm to spend months on. If you want to try doing the work on your own, you might want to start by looking for some tutorials on I2C. Once you understand how I2C works in general, and more specifically in linux, you should be able to understand how the beaglebone software could be adapted for Raspberry Pi. But, the hardware changes are a whole different matter. For that, you’ll need to understand electrical engineering and PCB layout.


Give me a chance to rephrase it again. In this GitHub link ( ) , in the installation explanation part you have put up " While this software is designed to be used on any embedded linux microprocessor, the current harware this codebase is being developed for is the Beagle Bone Black Wireless. If you are familiar with a Raspberry Pi, it is effectively the same thing with a few minor differences. " So does that mean I can install the same code for the RaspBerry Pi too? This is the question.


Ok chuck it, I can figure something out. Btw, I see that you have used CO2 sensors, Temperature and Humidity sensors. I actually have a few questions about it. 1) How are you controlling the CO2 level in the PFC after you determine the level of CO2 using the CO2 sensors?
2) What system do you have in place to control the temperature and humidity?

  1. I see using CV you guys are determining the growth of the plant, but is it necessary to determine the height of the plant? Or as you guys have implemented in the video i have seen you guys are considering the span o the plant and not the height, is only that much enough?


When you say “you guys”, I think you may not realize that the OpenAg team at MIT and the people who hang out here on the community forum are mostly different groups of people. The team at MIT is very busy with hardware and software design work, and they don’t come around here much. People here on the forum do a lot of projects that are inspired by MIT’s work, and sometimes use MIT’s hardware directly, when it is available. For example, the MVP food computer is a community designed and supported project.

You’re right that the PFC_EDU software repository mentions Raspberry Pi, and based on comments from Jake and Rob at MIT, they intended the software to be portable to other platforms. But, since the software is designed to control custom hardware (beaglebone connector & LED array are built into the same circuit board), the process of switching would require a new hardware revision among many other changes. If you want to do the work to make the necessary modifications for Raspberry Pi, the designs are all open source, so you’re free to do that.

As far as I can see, the only actuators on the PFC_EDU are lights and ventilation fans. There’s a lot of flexible control over the light spectrum, but I’m not sure if the fans can be switched. They might be always turned on. There’s no control over the CO2. There’s no control over temperature and humidity unless you count the heating and drying effect from waste heat generated by the LEDs. But, the LEDs have a big heat sink with fans, so heating in the plant chamber should be limited. You can verify all these things for yourself by looking at the designs in OpenAg’s electrical and mechanical repositories on github.

If you want to know about the MIT team’s plans for CV, the best place to look is the code and the github issues page for the repository you mentioned earlier. I don’t speak for the MIT team, and there’s not a lot of public information about what they’re doing with CV. Some of the software in the repository you mentioned seems to be intended for OpenAg’s Food Server container farm at MIT. They haven’t released much information about how that works, so your guess is as good as mine. Pictures on Caleb’s instagram are one source of information. There are also designs in the electrical and mechanical repositories on github.


I have more doubts, I went through the assembly instructions pdf. There it’s been mentioned that the foam padded sides help us to keep the temperature to the desired level. So how does it exactly do that I couldn’t understand. The DHT-11 sensor is being used to monitor the temperature and the humidity. So my assumption is whatever temperature is already there in the box after it’s completely built is maintained by the foam for the given growth period. Am I right?

/ 29


Oh ok ok now I got it.


The OpenAg designs are being developed and tested in a presumably rather cold warehouse in Massachusetts, USA (42 N latitude). The foam should help to retain some heat from the LEDs, and it could help to retain humidity if the surrounding air is dry. In a warm climate, insulation would probably not help much–you’d need a cooling mechanism.

[edit: When you said “foam”, I originally thought that you were using that word in an approximate way to refer to the acrylic panels of the PFC_EDU enclosure because you didn’t know the right term to use. Now I’m thinking maybe you were looking at the designs for the old “foam farm” activity and confusing that design with the design for the PFC_EDU. If you follow the links in OpenAg’s PFC_EDU announcement post, that should get you pointed in the right direction.]


So the PFC 3.0 is basically designed for leafy greens which require a warm climate?


I have a doubt, so is this for PFC 3.0 or the earlier versions?


Btw also,
There is one version of the PFC 3.0 brain software which divides the installation process into the following steps

  1. Purchase a BeagleBone Black

  2. Prepare a SD card image

  3. Flash image onto BBB

  4. User Setup

and another process under Contributing Software which says Getting Code in the same page that is.

What are the differences btw both these installation procedures?


That seems like a drastic over-simplification of the overall situation. If you care about getting a good understanding, I recommend that you keep reading and keep thinking about it. For example, have you considered the adjustable spectrum lighting?

That’s old stuff that’s left over from the PFC v2.x

I don’t have a simple answer for you. This stuff is all very new, and it’s quite complicated. I’m still studying it. If you want to understand the procedures, I recommend that you study the code and the scripts. If you learn something useful, perhaps you can post it here so others can benefit from what you’ve learned.


@wsnook Thanks for sticking with the thread, we appreciate your time and intent here!

@Vortex1 I’m with you… the required BeagleBone is way over spec for my planned builds and is simply not cost effective considering what I understand it’s used for IMO. That argument aside, it appears we’re on our own with regards to reverse engineering the new OpenAg Brain software. Hey, I’m thankful we even get those table scraps! At some point we’ll need a solution that suits wider application and I’m thinking what better place to start?


@Drew and @wsnook thank you all so much for your help. I have assembled my own team right now to develop the PFC using a Raspberry Pi, while I simultaneously work on learning as much as I can about agriculture. This project couldn’t have been started without the base you guys have provided. I will still need some help from all of you since I intend to develop a large scale version suited for my country India’s conditions. Can I please get your personal mail id’s so that I can keep in touch with you?


@Vortex1 I’ll be here, I’m not going anywhere :slight_smile: . You can always PM me here. When you get started please do drop us a line and tell us what you’re up to! Also, if you fork the codebase and begin work on the Rasp please let me know as I might decide to do the same. Thanks for contributing!


@Drew, sure I will keep in touch here itself, I just asked your personal mail id because in our previous conversation I was told that the core team members barely visit here, since they are really busy making new developments to the OpenAg project. Anyway, I will keep in touch. Thanks for your helping hand. :).


@Drew sure I will keep in touch with you here itself. Thank you for your helping hand :slight_smile:


Oh, I’m not part of that team :slight_smile: . I’m one of the random members of the community here to build cool horticulture tools and share my learnings. No worries, good luck and I look forward to your return!