What is pandar-40p?

Pandar40P Mobile Mapping Rig is a do it yourself project for building a walk around 3D mapping system. It combines a surplus Hesai Pandar40P lidar sensor, originally built for the automotive fleet market, with a jogging stroller, a laptop, and some custom electronics, to scan and map physical spaces in real time. The lidar is mounted on a tilted mast attached to the stroller, and paired with an inertial measurement unit, two cameras, and a GPS receiver. As the stroller is pushed around, the system uses real time lidar inertial SLAM, a technique for building a map while simultaneously tracking the device's own position, to generate colorized point clouds, 3D meshes, and Gaussian splats, which are all different ways of representing 3D scenes captured by the sensors. One notable finding documented in the project is that these fleet surplus lidar units use a type of automotive networking called 100BASE T1, instead of the standard Ethernet most computers use. Units that appear broken because they show no connection on a normal network card are not actually defective, they just need a roughly 100 dollar adapter to convert the signal so a laptop can read it. This detail alone could save someone from discarding a working sensor. The repository includes documentation for wiring the hardware safely, since the lidar spins as soon as power is connected and has no on off switch. It also provides scripts for setting up network hardware, discovering the lidar's IP address, and recording data. Firmware for a small microcontroller handles precise timestamping for the inertial sensor, and separate code publishes that data to the ROS 2 robotics software framework, alongside starting configuration files for the Hesai lidar driver and the FAST-LIO2 mapping algorithm. The project is still in an early, in progress state. According to the status list, one lidar unit has already been disassembled for parts and study, a media converter for the networking issue has been ordered but not yet tested, and steps like camera calibration, mounting the mast, and running an actual indoor mapping test have not been completed yet.