How to choose a brain for your robot?

1.2 Single-board Computers (SBC)

In the previous section, you have read about different vendors and different microcontrollers for building a robot. In this section, you are going to see about single-board computers. You know the microcontroller is a microcomputer, right? It has the same component that you can see on your computer. You can perform tasks such as sensor data processing, motor interfacing, etc using microcontrollers. The computation power of the microcontroller is very low compared to our personal computers.

If you want to handle a high-end sensor like a camera, Lidar, you may require more processing power and it can’t be done in a microcontroller platform. In this situation, you can either do the computation on a laptop or you can find another computing unit that can give the same performance as your laptop.

The computation modules that you use in robots should be compact and it should not take much power resource from the robot. It should not have much weight too. The kind of computers which satisfy the above requirements are called single-board computers. This makes the SBC as an ideal choice for robots. Even though it is not powerful in terms of computational power, but it can achieve most of the things that can be done with a PC.

Let’s have a look at some of the popular SBC used commonly in robots. You can also see the comparison of the board’s specifications. The specification of SBC is the same as your personal computer. It is having CPU, RAM, Storage, etc.

Here are the few SBC we are going to discuss in the upcoming section.

  1. Raspberry Pi

2. Odroid boards

3. Nvidia Jetson Nano

4. Nvidia Jetson TX2

5. Nvidia Jetson AGX Xavier

1.2.1 Raspberry Pi vs Odroid XU4

The Raspberry Pi series of boards and Odroid board series are entry-level SBC used as robot brain. In this section, you can see the comparison of the latest SBC from the Raspberry Pi foundation called Rpi 4 and the Odroid XU4 board from Hardkernel. You can see the image of these boards below, and the comparison of the important specification. This comparison can make the selection easier.

Raspberry Pi 4 and Odroid XU4
SpecificationsRaspberry Pi 4Odroid XU4
Processor with speedBroadcom BCM2711, Quad core Cortex-A72, 64 bit [email protected]1.5GHzSamsung Exynos5422 Cortex™-A15 2GHz and Cortex™-A7 Octa core CPUs
RAMUp to 4GB RAM (choices are 1 GB, 2 GB and 4 GB)2GB LPDDR3 RAM
USB Ports2 USB 3.0 and a USB 2.0 ports2 x USB 3.0 Host, 1 x USB 2.0 Host
ConnectivityGigabit Ethernet,
Dual-band Wi-Fi,
Bluetooth 5
Gigabit Ethernet,
Optional Wifi and Bluetooth
Display Port2 micro-HDMI ports (supports 4K displays)HDMI 1.4a for display
Power Consumption5V/3A5V/4A
PriceFrom $35/£34$49.00

Here is another detailed comparison between Raspberry Pi and Odroid.

1.2.2 Jetson Nano vs TX2 vs Xavier

The Raspberry Pi and Odroid were for entry-level robot brain. If you want to do heavy computational applications like robotic vision, deep learning and motion planning, you may need a high-end SBC like Nvidia Jetson series boards. Here is a comparison of 3 of the boards from Nvidia. This will help you to choose the right brain for your robot.

Nvidia Nano vs TX2 vs Xavier
SpecificationJetson NanoJetson TX2Jetson Xavier
Processor (ARM)4-core ARM A57 @ 1.43 GHz
4-core ARM Cortex-A57 @ 2 GHz, 2-core Denver2 @ 2 GHz
8-core ARM Carmel v.8.2 @ 2.26 GHz

GPU128-core Maxwell @ 921 MHz
256-core Pascal @ 1.3 GHz
512-core Volta @ 1.37 GHz

Memory4 GB LPDDR4, 25.6 GB/s
8 GB 128-bit LPDDR4, 58.3 GB/s
16 GB 256-bit LPDDR4, 137 GB/s

MicroSD32 GB eMMC 5.1
32 GB eMMC 5.1

(4x) USB 3.0 + Micro-USB 2.0
(1x) USB 3.0 + (1x) USB 2.0
(3x) USB 3.1 + (4x) USB 2.0

PCI-Express lanes
4 lanes PCIe Gen 2
5 lanes PCIe Gen 2
16 lanes PCIe Gen 4

5W / 10W
7.5W / 15W
10W / 15W / 30W
Price99 USD412 USD712 USD

Lentin Joseph

View posts by Lentin Joseph
Roboticist | Author of 8 robotics books | TEDx speaker

Leave a Reply