To address the challenges of multi-motor coordinated control in soft robot applications, such as high difficulty, poor versatility, and poor coordination, a Controller Area Network (CAN) bus motion controller for soft robots based on ARM and FPGA was designed. This CAN bus motion controller for soft robots was implemented using an STM32F407 development board, featuring an ARM Cortex-M4 core, and an AX7102 FPGA development board. The design primarily covers the system architecture, hardware design, and software design. The controller leverages the STM32 as the control core and the high-speed processing capabilities of the FPGA to execute control algorithms, while CAN bus technology is employed for communication with the host computer. Experimental operation has demonstrated that this controller meets the predefined requirements.

Soft robots differ from traditional rigid robots, offering numerous advantages, including bio-flexibility. They can perform flexible obstacle avoidance or compliant grasping operations by peristalsis, crawling, and twisting through confined spaces, demonstrating broad application prospects in unstructured environments [1]. Consequently, there is a need for a motion controller specifically adapted for soft robot motion control. To date, general-purpose motion controllers can be structurally categorized into the following three main types [2-3]:

(1) Standard Bus Motion Controllers. These motion controllers are mostly based on standard computer buses, operating independently of the computer with an open architecture. They typically use Digital Signal Processor (DSP) chips or microcontroller chips as their processors, capable of implementing functions such as motion planning