Your question actually includes two aspects: design and production. Making is easier than designing, because copying a robot can also be said to be making. The required knowledge is as follows (refer to the course planning of engineering university): structure (related to mechanical specialty): mechanical principle (basic kinematics and mechanical principle of required mechanism), mechanical design (understanding the use and basic matching relationship of various parts), basic assembly method (many structures can be designed, but cannot be assembled), engineering drawing (at least two-dimensional drawings should be made when handing them over to others for processing), tolerance calculation basis (knowing which dimensions need high precision when processing), Engineering material foundation (it is enough to know the characteristics of various commonly used materials and some commonly used composite materials) and processing technology foundation (know how to process various parts, mainly cold working). Hardware circuit (related to electronics): basic principle of circuit, analog circuit, digital circuit (basic control circuit: power supply, logic circuit, etc.). ), detection technology (understand the principle, indicators and usage of common sensors). Software (related to computer specialty): programming language (C language or assembly language), operating system foundation (helpful to design more complex software architecture and learn controllers such as single chip microcomputer and DSP), common algorithms and data structures (helpful to design reasonable, efficient and creative robot control algorithms), computer network foundation (such as future multi-robot communication), software engineering (knowing how to design and maintain software) control (and automatic control specialty) Electromechanical (know how to control the motor rotation), automatic control principle (let the motor rotate according to your desired speed and target position, such as the classic PID algorithm), and some knowledge of signal processing (such as filtering). Mathematics (related to engineering majors): calculation methods (some algorithms used in actual calculation), linear algebra (or advanced algebra and matrix theory) are necessary tools for kinematics and dynamics calculation of multi-joint robots in the future. Accounting (related to business major): cost analysis, budget and other concepts, at least you need to know how much it will cost to make this robot (generally, processing is the main thing, followed by drawing circuit boards, and if you do it yourself, the accounting cost of software development can be ignored). Specific to the technical aspects that may need to be learned, we can dabble in the following aspects: structure: learning a three-dimensional modeling tool (SolidWorks or Pro E, UG, etc. ), you can make two-dimensional drawings; It is best to do some finite element analysis (it will help you design a reliable structure). Hardware circuit: Electronic design tools (Protel, PowerPCB, etc.) can be used. ), can draw schematic diagram and PCB, a circuit simulation tool (EDA software: such as Proteus), single chip microcomputer (5 1, AVR, Freescale, ARM, etc. ) or CPLD/FPGA, your driving and control algorithm should be realized on it. Software: Learn an IDE (such as Keil, ICC, Realview, etc.). ) and understand the editing, compiling and debugging methods; Learn to use uC/OS or other RT OS. If you want to run an operating system, it is easy to transplant. If classical control wants to design its own robot, I'm afraid it needs to have its own reliable ideas on the basis of certain "manufacturing", which can be deepened from many disciplines, to give a few examples: mechanism: learn more about the principles of mechanism and accumulate more experience in mechanical design; Digital image processing: target detection, image recognition, stereo vision, etc. Control theory: filtering algorithm, system identification, adaptive control, fuzzy control, etc. Bionics: Many novel robots are based on bionics. In addition, you can also pay attention to international robot-related papers (such as Chinese version of robotics) or conferences (such as ICRA and IROS). ) and follow up the recent research on robotics in famous universities (such as CMU and MIT). ), it will definitely broaden your horizons.