CUSTOM BUILT HANDHELD DEVICE
I created a custom tablet device capable of running very complex SLAM algorithms in real-time consuming streams of data from depth sensors to create 3D map of the physical environment. I assembled the components, wrote software to integrate open source SLAM solution with depth sensor data on a Linux (Ubuntu) baed platform. Here are the components I used to build the device.
​
Nvidia Jetson TX2
Main compute board SOC (System on a chip). It has CPU and GPU cores and is worlds smallest super computer
J120 Carrier Board
Carrier board to mount the main SOC on. It exposes various IO connectors to interface peripherals with the system like HDMI, USB 3.0, Power module for battery connection and GPIO
7 inch LCD Screen
Display for the table device. It is a 7 inch multi-touch display.
Intel RealSense D435 Stereo Camera
Stereo depth sensing camera with two IR sensors, one RGB sensor and an IR blaster
Assembly
These components were assembled together along with a rechargeable battery to create a very powerful 3D sensing device. Main OS was Ubuntu 16.04 LTS and all software was written on this platform. The diagram above shows the assembly of main components