-
Notifications
You must be signed in to change notification settings - Fork 27
Parts List
I estimate that the absolute minimum cost is going to be $1,500. Here I have included the minimum parts list with prices where I can.
This is the minimum cheapest build that will function. Upgrades and improvements will be on the next page.
I will include links where I can find them to possible sources, although some small items I obtained locally from Radio Shack and/or Ace Hardware.
The majority of the robot I use is just an Arlo Robotic Platform from Parallax. Parallax now sells a complete kit:
-
Arlo Complete Robot System $1,195.00
NOTE: This kit uses a new motor control board called the "DHB-10 Dual H-Bridge 10 Amp Motor Controller" instead of the HB-25s that were previously used. My robot has HB25s. I am committed to supporting the DHB-10, but since I do not have one I cannot guarantee that it works yet. If you have trouble with it working with my code contact me and we can work together on sorting out the differences. I am also confident that Parallax will help us on that.
ROS needs a "LIDAR" or "Laser Scanner", but they are expensive, so the TurtleBot, and our robot, uses a 3D camera sensor instead. This can be either an XBox 360 Kinect or an Asus XTION PRO LIVE/B/U Depth Sensor or maybe an Orbbec Astra
NOTE: The XBox ONE sensor will not work. Although I've seen rumors online that it may be supported? Do some research.
I suggest looking on Ebay for a Kinect sensor.
-
Kinect 360 Sensor on Ebay $40 (guess)
- You will need a splitter for power if the sensor you get does not include one:
-
Xbox 360 Kinect AC Adapter/ Power Supply $11
- You can cut the AC adapter off of this and wire it to 12 volts.
OR
-
Asus XTION PRO LIVE/B/U USB 2.0 and USB 3.0 RGB and Depth Sensor $169.99
NOTE: The Asus MUST be the "LIVE" version!
- The main benefits of the Asus over the Kinect are:
- Smaller
- Runs off of USB power
- These MAY BE out of production, but they seem to show up again periodically. Just keep checking back.
- The main benefits of the Asus over the Kinect are:
OR
-
Orbbec Astra $149
- TurtleBot is experimenting with using the Orbbec Astra 3D camera: http://discourse.ros.org/t/new-depth-sensor-orbbec-astra-call-for-testing/252
- I do not have one of these, and the support is still experimental, but I think this camera is the future direction for ROS. I would recommend it.
- Please check out this link and the links on it for more details: http://discourse.ros.org/t/new-depth-sensor-orbbec-astra-call-for-testing/252
- I am not sure which "version" of the camera is best. I think you do NOT want the "S" version, but do some research and check the above link for more details.
ROS Runs on Ubuntu on a PC. It doesn't work on Android or Raspberry Pi. Maybe someday, but right now that is not going to accomplish the task.
The TurtleBot comes with an ASUS “Netbook” with an Atom N525 dual-core 1.8GHz CPU. While this works, I think it should be considered the absolute minimum for a usable laptop for ArloBot.
-
HP EliteBook 2760p from Ebay $250
- This is what I bought. Obviously buying a laptop on Ebay is a crap shoot. Fortunately if it is beat up, or the keyboard or touch pad are broken that is ok.
If you have an old laptop, or can borrow one, that is the best way to get started. Once you figure out what you want, then keep an eye on Amazon and/or Ebay for one you can buy.
Remember that it is easy to dual boot Ubuntu with Windows, so you can always use a laptop that has to perform other duties to also run on the robot.
-
SainSmart USB Eight Channel Relay Board
- This isn't required, but its use is built into my code and it will make your life so much easier. Without it you will be fussing with turning the robot's motors on and off at just the right time. <- This may not be true of the new DHB-10? With it the motor startup will be automatic and you can bring up the robot remotely.
- There is also a 4 port version that is smaller and a little cheaper. I haven't tried it, and the code will have to be modified to work with it, but I can help you with that and then we can make it automatic for future users.