Installing ROS on a Raspberry Pi

As a lover of technology, I tend to accumulate bits and pieces of interesting devices. Usually, these are purchased for use on unrelated projects, and on occasion, I have the opportunity to bring them together into a single project in a previously unanticipated way. Such is the case with my Arduino and Raspberry Pi. Both are interesting microcomputers with their own strengths and weaknesses, so it was when I learned that they could be made to work together with the help of Robot Operating System, I had to give it a shot.

My raspberry pi

ROS is an open-sourced project that is dedicated to providing a framework of libraries for performing common tasks under the general heading of robotics. It also includes drivers that allow you to easily interface with common hardware. The core of ROS is a reactor model of observables and observers that send messages to one another, typically over a serial connection, allowing any number of controllers to interface with one another and form a unified whole.

The rosserial_arduino library is a project that allows ROS on a Raspberry Pi (or other *nix device) to interface with an Arduino over a USB serial connection, thereby combining the computing power and versatility of a Linux-based microcomputer with the IO capabilities of an Arduino.

What You’ll Need to Get Started

Installing Raspbian on the Pi

If your Pi already has an operating system on it, you can probably skip this step. If, however, it’s straight out of the box, you’ll need to install the Raspbian distribution.

As of this writing, the latest version of Raspbian is Jesse, released in September of 2015. I wasn’t able to get ROS working with this version, and backed down to the Wheezy release from May of 2015 instead. To install the operating system, I did the following:

  1. Download the Raspbian Wheezy image via a bittorrent client.
  2. When the download is complete, follow these instructions to copy the image file to your MicroSD card.
  3. Unmount the card, insert it into your Pi, and hook up the power. Your device should boot into a command prompt. From here, you can run raspi-config to customize the installation, or get right to installing ROS.

Once the installation is complete, be sure to check for updates:

pi@raspberrypi ~ $ sudo apt-get update
pi@raspberrypi ~ $ sudo apt-get upgrade

An up to date system is a safe system.

SSH

Once your Pi has an operating system, you can switch to interacting with it via SSH. My TV is the only “monitor” in my house that has an HDMI input on it, so SSH works much better for me.

Make sure that sshd is running on your Pi:

pi@raspberrypi ~ $ sudo service sshd status
● ssh.service - OpenBSD Secure Shell server
 Loaded: loaded (/lib/systemd/system/ssh.service; enabled)
 Active: active (running) since Thu 2015-10-08 12:17:06 UTC; 4 days ago
 Main PID: 506 (sshd)
 CGroup: /system.slice/ssh.service
 └─506 /usr/sbin/sshd -D

If everything is working, you should see the text active (running) in the result. Once we know that an ssh server is running, we can check our ip address with the ifconfig command. The output should look something like this:

pi@raspberrypi ~ $ ifconfig
eth0      Link encap:Ethernet  HWaddr b8:27:eb:b9:49:cd  
          inet addr:192.168.0.109  Bcast:192.168.0.255  Mask:255.255.255.0
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:5150 errors:0 dropped:51 overruns:0 frame:0
          TX packets:565 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:552488 (539.5 KiB)  TX bytes:60766 (59.3 KiB)

lo        Link encap:Local Loopback  
          inet addr:127.0.0.1  Mask:255.0.0.0
          UP LOOPBACK RUNNING  MTU:65536  Metric:1
          RX packets:8 errors:0 dropped:0 overruns:0 frame:0
          TX packets:8 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0 
          RX bytes:1104 (1.0 KiB)  TX bytes:1104 (1.0 KiB)

If your Pi is connected to a LAN cable, you’ll want to look at the eth0 section. If it’s connected to WiFi, look for a wlan0 section. Both sections should have an inet addr field whose value starts with a 192.168.x.x address. In my case, it’s 192.168.0.109. From a terminal on my computer, I can connect with:

jfritz@IDEAPAD-UBUNTU:~$ ssh pi@192.168.0.109

When prompted to accept the Pi’s RSA key, I do, and when prompted for a password, I enter the default password raspberry. If you intend to leave the Pi connected to your network for long periods of time, you should change this password or add key-based authentication to the system.

If you have problems getting connected, check out the official instructions on the Raspberry Pi website.

Installing ROS on the Pi

As of this writing, the most recent version of ROS is Indigo, released in July of 2014. To get it running on the Pi, you’ll want to follow the official ROSberryPi installation instructions on the ROS website.

While following these instructions, I had a few false starts. It’s important to read the instructions carefully, as they’re fairly generic, and can be used to install different configurations of ROS on different versions of Raspbian. I found that the instructions for the ros_conn configuration worked best on Raspbian’s Wheezy release.

The  trickiest part of the instructions is section 2.2 Resolve Dependencies. It took me a couple of reads to realize that if you’re installing ROS Indigo’s ros_conn configuration on Raspbian Wheezy, you only need to compile two packages from source: libconsole-bridge-dev and liblz4-dev. Installing any other packages at this step just costs you time, and may introduce problems down the road.

I also found that the install process went much smoother when the Pi was connected to a LAN rather than WiFi. The WiFi signal in my house is relatively weak, and the Realtek #814B is really cheap, so downloading a lot of files while maintaining an SSH connection is a big ask.

Once the installation is complete, open up your ~/.bashrc file, and add two lines to the end:

# export ROS environment variables
source /opt/ros/indigo/setup.bash

This will make sure that the appropriate environment variables are set to interact with ROS on every startup. You can check that it worked by rebooting your Pi and running

pi@raspberrypi ~ $ printenv | grep ROS
ROS_ROOT=/opt/ros/indigo/share/ros
ROS_PACKAGE_PATH=/opt/ros/indigo/share:/opt/ros/indigo/stacks
ROS_MASTER_URI=http://localhost:11311
ROS_DISTRO=indigo
ROS_ETC_DIR=/opt/ros/indigo/etc/ros

If you see all of the ROS_* environment variables print out, then everything is set up and ready to go. Now it’s time to start on some tutorials.

Eventually, I want to get the Raspberry Pi communicating with the Arduino, and use the latter as a sensor platform and motor controller for some kind of a robot. For now, I need to find my way around ROS.

1 Comment

Filed under RaspberryPi, ROS

One Response to Installing ROS on a Raspberry Pi

  1. Pingback: Installing ROS on a Raspberry Pi | The Linux Experiment

Leave a Reply

Your email address will not be published. Required fields are marked *