1.2 Mini Computers
The boards that we have seen in the above sections are based on ARM architecture. If you want to use an x86, x86_64 based processor in your robot, you may choose a mini-computer. So what is a mini-computer? It’s the same as your laptop and desktop computer but with a miniature in size. So what is the advantage of using it in your robot? The x86 and x86_64 are the popular architecture used in PC, so most of the software libraries are fully compatible with that architecture. In most of the ARM processor-based boards, there will be some compatibility issues with existing software. It is a cumbersome task to solve those issues.
In order to minimize software compatibility issues, you can directly use the mini-PC. One of the commonly using mini-PC in robots is Intel NUC (Next Unit of Computing). One of the advantages of using mini-PC is, it will support all the popular operating systems, but that is not the case of SBC boards. Most of the board is only compatible with custom Linux configuration and some support android too.
There are different types of mini-PC available from different vendors. In this section, I will be mainly discussing three different types of Intel NUC mini PC.
The form factor and shape of NUC will look like in the following figure.
|Specification||Intel NUC: NUC5CPYH||Intel NUC: NUC7i3BNK||Intel NUC: NUC8i7BEH|
|Processor||Intel® Celeron® Processor N3060 (2M Cache, up to 2.48 GHz)||Intel® NUC Kit with 7th Generation Intel® Core™ Processors, Intel® Core™ i3-7100U Processor||Intel Core i7-8559U processor (2.7 GHz – 4.5 GHz, Quad Core, 8MB Cache, 28W TDP)|
|Max RAM support||8 GB,DDR3L-1333/1600 1.35V SO-DIMM||32 GB, DDR4-2133 1.2V SO-DIMM||Two DDR4 SO-DIMM sockets (up to 32 GB, 2400 MHz), 1.2V|
|Intel® HD Graphics 620, HDMI 2.0a; USB-C (DP1.2)||Iris Plus Graphics 655|
|PCI Express Configurations||M.2 slot with PCIe X1 lane||M.2 slot with PCIe x4 lanes||M.2 slot with PCIe X4 lanes|
|Connectivity||Intel® Wireless-AC 3165 + Bluetooth 4.2, Ethernet||Intel® Wireless-AC 8265 + Bluetooth 4.2, Ethernet||Intel® Wireless-AC 9560 + Bluetooth 5.0, eTHER|
|Input voltage||12-19 VDC||12-19 VDC||12-19 VDC|
|Supported OS||Windows, Linux||Windows, Linux||Windows, Linux|
|Price||$ 170||$ 450||$538.20 – $540.50|
|Weight||~ 530 g||~ 530 g||~ 530 g|
In the above table, you can see three different NUC with different configurations. Depending on your robotic applications, you can choose any of this mini-computer. Here is the detailed comparison of all Intel NUC boards.
Other than Intel NUC, there are other vendors manufacturing mini-pcs. Here is a detailed list of popular mini-pc and their vendors.
If you are building robots like self-driving cars, you may need high-end computers in order to perform its computation. Here are a few examples of high-end computing platforms we can use for complex robots.
For example, the famous self-driving car company, Waymo is using the Intel Xeon servers, Intel Arria 10 FPGA board and GPU from Nvidia. Here are few high-end computing units used for complex computation
Here is an introduction video of Nvidia Drive AGX board
1.5 Industrial PLC/SCADA system
The Programmable Logic Controller (PLC) is an industrial digital computer which has been using for control of manufacturing processes such as assembly lines, robotics device, etc. Supervisory control and data acquisition (SCADA) is a control system architecture comprising computers, network data communications and graphical user interfaces (GUI) for supervisory management. The PLC and SCADA systems are commonly used as a brain in industrial robots.
Given below an example PLC module from Siemens