Vending Machine Blueprint
Contents
Overview
The Vending Machine Blueprint provides support for building a PoC smart vending machine with components showing the capabilities of the Oniro Project build infrastructure to easily build an end-to-end solution using multiple operating systems cooperating inside a product.
Hardware Setup - Bill of Materials
The Vending Machine blueprint provides full support for two complete setups. One based on Raspberry Pi 4B, an ARMv8 target, and one based on SECO B68, an X86 board. Regardless of the target board chosen, you will need additional components, connectors, and power supply. Besides the target board, the other main differences in terms of hardware are the power supply and the HDMI adapter. The latter is needed because the boards have different video input connectors (miniDP vs micro-HDMI) while the display comes with a standard HDMI input connector and cable. Please find below the full bill of materials for the two supported setups:
Raspberry Pi 4B-based setup BoM
1 x 7inch HDMI LCD Display, Waveshare, Rev3.1 or Rev2.1 (comes with USB and HDMI cable)
4 x multi-coloured male to female DuPont wires (preferable colours: black, red, yellow and green)
1 x microSD card (>=8GB size)
1 x Raspberry Pi 4B board (>=2GB RAM)
1 x HDMI to micro-HDMI adapter
SECO B68-based setup BoM
1 x 7inch HDMI LCD Display, Waveshare, Rev3.1 or Rev2.1 (comes with USB and HDMI cable)
4 x multi-coloured male to female DuPont wires (preferable colours: black, red, yellow and green)
1 x microSD card (>=8GB size)
1 x SECO B68 board
1 x HDMI to miniDP adapter
1 x 19V power supply - depending on the board variant, you might need a barrel or an industrial screw connector
Building the Blueprint Hardware Setup
Once we have all the required parts for this blueprint, we can proceed to build it. Due to differences in connectors and power, there will be a set of differences between the two setups but in general, the steps are:
connect the display’s HDMI input to the board through the appropriate HDMI adapter
connect the display’s USB cable to the USB connector on the board
connect the RGB Matrix to the right I2C pins on the board using dupont wires
Note
Note that the RGB Matrix does not come with a soldered pin header so some soldering work will be required.
Raspberry Pi 4B-based Hardware Setup
When building a Raspberry Pi 4B-based hardware setup you will need to:
connect the display’s HDMI input through a micro-HDMI adapter to the connector labeled HDMI0 on the Raspberry Pi 4B board (see the image above)
connect the RGB LED Matrix to the Raspberry Pi 4B I2C interface:
SECO B68-based Hardware Setup
When building a SECO B68-based hardware setup you will need to:
connect the display’s HDMI input through a miniDP adapter to the connector labeled CN20 on the SECO B68 board (see the image above)
connect the RGB LED Matrix to the SECO B68 I2C 4 port on the CN19 connector:
The CN20 connector on the board can be seen here:
Fetch Blueprint Sources and Build
Fetch the Build Metadata
The final piece remaining after building the hardware setup is fetching the software sources and building the OS.
For fetching or updating the build metadata, follow the blueprints workspace documentation. This will bring everything you need on your host to start building an image for this blueprint.
Note
Mind the CHECKOUT_DIR variable in the workspace documentation as you will need them below.
Build the Oniro Image for the Vending Machine Blueprint
Once you have a workspace initialised as per the instructions above, you are ready to build the OS. Firstly, you will be initialising a build:
$ TEMPLATECONF=../oniro/flavours/linux . ./oe-core/oe-init-build-env build-oniro-vending-machine
This will setup a new build environment in the build-oniro-vending-machine directory (or reuse it if it already exists).
Add the meta-oniro-blueprints layer to the build. This is only needed once after initializing a new build:
$ bitbake-layers add-layer $CHECKOUT_DIR/meta-homeassistant
$ bitbake-layers add-layer $CHECKOUT_DIR/meta-oniro-blueprints
Once the build environment is set we can start building the image. The building command will be slightly different depending on the target board.
For the Raspberry Pi 4B-based setup:
$ DISTRO=oniro-linux-blueprint-vending-machine MACHINE=raspberrypi4-64 bitbake blueprint-vending-machine-image
For the SECO b68-based setup:
$ DISTRO=oniro-linux-blueprint-vending-machine MACHINE=seco-intel-b68 bitbake blueprint-vending-machine-image
Once the build is finished, the images are available for being written on a microSD card.
Flashing the Blueprint Image
First of all, make sure you have plugged in an microSD card via an microSD card reader attached to your host. Once that is done, note its associated device node. You can find that using udev, dmesg or various other tools. Once the device not associated to the microSD card is known, proceed to flash the built image.
Warning
The commands below assume that the device node associated with the microSD card is provided via the DEVICE environment variable. Make sure you have the correct one set before running the commands below to avoid risking data loss.
The image to flash is a full disk image so the DEVICE variable needs to point to the entire block block device node and not to a specific partition.
For the Raspberry Pi 4B-based setup:
$ sudo bmaptool copy ./tmp/deploy/images/raspberrypi4-64/blueprint-vending-machine-image-raspberrypi4-64.wic.bz2 "$DEVICE"
For the SECO b68-based setup:
$ sudo bmaptool copy ./tmp/deploy/images/seco-intel-b68/blueprint-vending-machine-image-seco-intel-b68.wic "$DEVICE"
Running the Vending Machine Blueprint Image
At this point, we have all the pieces we need to run the entire blueprint. Plug in the microSD card on which you flashed the OS into the board’s microSD card slot. That brings us to the last step: booting the board by the attaching power source. On a Raspberry Pi 4B, that would be a USB-C power supply while on SECO B68, it would be a 19V power supply connected to the power barrel or the industrial screw connector (depending on the board’s variant).
Once the boot process completes, the vending machine UI application will take over the screen and be ready to interact with the RGB LED matrix for item selection and delivery simulation.
