CEB Automation Strategy

Currently, we are still working on getting our first order of The Liberator high performance, open source, Compressed Earth Brick (CEB) press out the door. We are now upgrading the automatic controls.

We have published a technical paper on the automation problem statement. We are managing the project at the Open+Pario project management site, where you can download a copy of this paper under the Documents tab. The paper outlines the technical issues surrounding effective automation, based on our previous results. It provides the necessary background if you want to collaborate on the project or build upon it – in the name of open source development. Here is the abstract:

Abstract: There are several considerations for the successful implementation of automatic controls on The Liberator*, the world’s first, high performance, open source, Compressed Earth Brick (CEB) press. Considerations include: (1), brick production work flow design; (2), simplicity of control logic; (3), brick thickness adjustability and uniformity; (4), modular, lifetime design; (5), performance optimization; (6), cost, (7), interchangeability of hydraulic power units1; (8) open source standards; and (9), simple user interface. This technical paper documents these issues for the development team and the greater community as part of the open source process. The greater context for this work is promoting the creation of post-scarcity, resilient communities. This paper promotes the greater context by contributing to the solution of one of the most basic needs of humanity – housing – under the assumption that earth construction is the most robust and most widely-used method of housing worldwide2. Revisions of this work are found at Open+Pario3.

1The Liberator is designed to be powered by any external, hydraulic power source, such as Power Cube* or another tractor or skid loader*. This is part of the modular design strategy of OSE and the Global Village Construction Set* (GVCS).

2See Earth Construction*. Regarding the extent of use of earth construction, the industrialized world is presently lagging behind the developing world.

*Footnotes labeled with an asterisk (*) refer to pages on the OSE wiki ( with the corresponding title. For example, this particular link for The Liberator is Subsequent words marked with an asterisk are documented at a wiki page with the corresponding name.



  1. LucasG

    Hi, I just directed folks from this way.

    Their story:

    My comment (hope it’s not considered spam):
    LucasG – Subject: high-end hardware replication too folks are building a CNC torch table with open design for self-replication. They already have a unix-tractor: a box on wheels with many implements. And a dirt-to-brick machine: capable of outputting 10 bricks per minute. And it’s even more ambitious than that.
    The funding strategy is twofold: true fans (please consider helping) + selling products (they’ve started already).

  2. Pawel Sroczynski

    reporting a bug in fils, my AutoCAD 2007 is saying: “Invalid or incomplete DXF input — drawing discarded.” – drawer.dxf

    Importing to SketchUP works fine.

  3. Marcin

    Pawel, Thanks for letting us know. Have you tried trashing AutoCAD and installing a free version of QCad community edition instead on Linux? Hahaha.

  4. Marcin

    From Ralf:

    Question 1: Can I do 7 sensor inputs to Arduino if it has only 6 input pins? Is this easy or not worth the effort? After thinking about the controls strategy, it would be easier to do this than to do a mechanical switch instead of a 7th input.

    You can use transfer gates and a shift register to switch the analog input to one of the arduino inputs. In that way you can connect multiple
    sensors to one input. Depends on how fast you need the inputs. What sensors are you using?

    That’s also what I’m planning to use for the solar concentrator prototype electronics: Just use some photo-resistors connected to a series of transfer gates and connect them through to the arduino using a shift register. In this way you only need one Arduino for *many* individual mirrors.

    Note that afaik the Arduino only has a single AD converter that is multiplexed among the different input pins. So you might need external AD converters if speed matters. The Dallas/Maxim 1Wire bus has interesting hardware that als reduces wiring (you only have ground and the 1wire control pin and put all the sensors or a/d converters on the same bus).

    Question 2: How feasible is it to have RFID or wireless sensors? That would simplify the wiring greatly, but it will make the core circuit more complicated.

    I don’t think there are RFID sensors. For wireless sensors: There may be some Zigbee sensors out there and I think there’s a Zigbee module for the arduino. I’ve not used it myself, though. Wireless sensors may pose a safety risk if a critical control loop is closed via wireless (in case of transmission problems the control loop wouldn’t work).


  5. Marcin

    From Leo on the same questions as in last comment:

    Q1. It’s usually possible to multilplex inputs. If the 7th input is worth having I’d be inclined to switch to a microcontroller with more inputs, though, since that’s simplest. Perhaps an Arduino Mega or a Sanguino.

    Q2: How feasible is it to have RFID or wireless sensors? That would simplify the wiring greatly, but it will make the core circuit more complicated.

    I’d recommend against it. Wireless tends to be less robust, and adds cost and complexity to the hardware and more complexity to the software. Also, you still need power wires, so you don’t win as much as you’d hope.

  6. Pawel Sroczynski

    Marcin: heheh 😉 no I didn’t 😉 QCAD is far not enough for me, unfortunately :/