Blogs

Bambu Lab Recently I needed to make my own mounting parts because I needed to install an Intel Realsense D435i for Vision SLAM on my Robomaster EP, so I decided to get my hands dirty and use 3D printing to do it. Although I have an overlord pro on hand that I purchased five years ago, due to the age of the machine, I need to manually optimize and tweak the performance of this printer, and considering the cost of time, I decided to purchase a new 3D printer for the printing of accessories, this overlord pro will be repaired later when I have time. ...

MPICH Raspberry Pi Cluster The earliest project to run MPICH on a Raspberry Pi cluster would be Raspberry Pi at Southampton, in which professor Simon Cox and his son together to build an MPI cluster consisting of 64 Raspberry Constructs and Legos. This project is already 10 years old, and at that time the raspberry pi 1 B+ was used, and due to the limitations of the performance of the 1st generation, the whole cluster could only be used for MPI learning, which is still under-utilized in high-performance computing. However, now raspberry pi 4 is still very weak compared to Intel servers, but it still has great potential to run MPICH related computing projects compared to the 1st generation. First of all, power consumption, 64 raspberry pi 4 CPUs running at full load consume about 400 watts, and secondly, the ARM ecosystem is now very well developed compared to a decade ago, and a lot of programs have been ported to the ARM64 platform, with a variety of tools at your fingertips. ...

Problem The system of raspberry pi zero w is currently raspberry pi official os image is armv6hf debian 11, and ROS2 official support only ubuntu debian aarch64, not support arm32, so to use ROS2 on raspberry pi zero w must be built from source. So the problem is that if you compile directly on zero w, obviously because of the hardware (cpu: 1 core, memory: 512M), the compilation will be extremely slow, I also tried to use armhf cross-compiler toolchain to compile the ROS source code on aarch64 system, but there will be a problem of missing system libraries. At the same time, I also consider the consistency of the subsequent development environment and zero w’s original system environment, I decided to use zero w’s native debian 11 armhf system as rootfs, make a docker image, and then run on aarch64 systems (such as raspberry pi 4, jetson agx), so that the compiled ros2 can be run on zero w directly. This way, the compiled ros2 can be run on zero w, and the runtime and compilation environments are consistent, which will be very convenient for subsequent tuning and debugging. ...

reference wolframalpha-as-the-way-to-bring-computational-knowledge-superpowers-to-chatgpt chatgpt-gets-its-wolfram-superpowers what-is-chatgpt-doing-and-why-does-it-work Next-Generation Human-Robot Interaction with ChatGPT and ROS Dolly v2.0 Principle building The recent hot chatgpt, advancing generative AI to a new level, but currently see most people apply it to chat tools, and even some people use it as Wikipedia, but I think since it is generative, why not use it to do the interaction between human and robot, we use a specific defined natural language to describe the operation we need to let the robot do. chatgpt translates our natural language into a structured language similar to program source code by understanding it, and then inputs it into the robot’s system, and the robot operating system will follow the input structured language to complete the corresponding operation. So I was thinking how to combine RoboMaster EP with LLM language class model to accomplish human-robot interaction. API applications like chatgpt have a disadvantage because their models run in the cloud and are restricted to certain countries and regions, so commercial industrial LLM applications are basically ruled out and I wanted to be able to run the models locally and train and optimize them by myself. At this time Databricks released Dolly v2.0 model, the dataset and the training model are all open source and can be applied to the commercial domain, after reading the related introduction I decided to use dolly v2.0 as my first choice for the local LLM model. ...

DJI RoboMaster EP At the beginning of the year I planned to spend six months researching autonomous driving, and computer graphics, so I started researching hardware for developing autonomous driving systems. The first thing I considered was the jetbot that was open when nvidia released the jetson nano platform in 2019, which is basically a toy car equipped with a 3D printed body, two small wheels, and a jetson nano developer kit, a camera, and official project address, considering the cost needed to build the entire car hardware platform and the playability, I decided not to consider a toy-like solution for the car robot, and finally I aimed at the DJI robomaster EP because It opened up the complete SDK, and the whole robot platform was designed for educational competitions, which is very rich in playability and does not require too much thinking about the underlying hardware assembly itself. This allowed me to take a page from jetbot’s autopilot design and port it to the robomaster EP to create the most original autopilot development platform. ...