Clay McPherson "Rebrains" a 1990s Nomadic N150 Robot with an Espressif ESP32, Teensy Pairing
A brain transplant has got this 16-sonar robot back up and running — with future upgrades planned to add modern SLAM capabilities.
Engineer Clay McPherson is working on bringing a bit of 1990s robot history back to life, performing a "brain transplant" on a Nomadic Technologies N150 robot — with a vision to eventually run the Robot Operating System (ROS) on the gadget.
"This is a Nomadic Technologies N150 that I got at auction for dirt cheap in a non-functional state," McPherson says of the chunky robot chassis he's been investigating of late. "It was manufactured some time around 1998-99. It uses 16 ultrasonic rangefinders, physical bumper sensors, and dead reckoning to navigate. I’ve rebrained it with modern microcontrollers and am working on implementing ROS to use some modern SLAM [Simultaneous Localization and Mapping] algorithms."
Designed primarily for education and research, the Nomadic Technologies N-series are a relatively unusual find today — which makes getting one back up and running again a challenge, thanks to a lack of documentation. While McPherson estimates that "presumably a couple hundred of them are floating around somewhere in old university labs," he's been going hands-on in his efforts to build the system back into working condition.
After opening the robot and checking the motors and sensors, McPherson set about performing a "brain transplant:" the removal of the original motherboard, now replaced by an Espressif ESP32 and a Teensy microcontroller pair. These operate the existing ultrasonic array and control the drive motors, with the ESP32 using its Wi-Fi radio to communicate with a graphical user interface for remote control — replacing a rather simpler gamepad direct-drive interface installed to prove the motor systems operational.
McPherson's overall goal, though, expands beyond basic control: adding a single-board computer to the mix, capable of hosting the Robot Operating System (ROS) and performing simultaneous localization and mapping (SLAM) — turning the feed from those 16 ultrasonic sensors into a point cloud representation of its environment for true autonomous pathfinding and map-making.
More information is available in McPherson's Reddit thread, and in the videos above; updates will be posted to his YouTube channel, Clay Builds. Initial source code has been published to GitHub under an unspecified license.