如果高中阶段就开始真正做产品，而不是只做作业，会发生什么？
What would happen if high school students started building real products, instead of only completing assignments?
我们正在尝试建立的，不是一门兴趣课，也不是普通的机器人社团，而是一间真正具备研究气质与创业精神的实验室——原点未来工程研究所 (Origin Future Engineering Lab）。
What we are trying to create is not simply an interest-based class, nor a typical robotics club. Instead, it is a laboratory that carries both a research spirit and an entrepreneurial mindset — the Origin Future Engineering Lab.
在这里，一切从最基础开始。
Here, everything begins with the fundamentals.学生不会直接拿到一台“已经调好”的机器狗，然后学习如何遥控它。相反，他们会先把它拆开。看清楚每一块结构板是如何连接的，每一个舵机的扭矩是否足以支撑身体重量，每一颗螺丝在受力时承担了什么角色。机器不是一个黑盒子，而是一个等待被理解、被改造、被重新设计的系统。
Students will not be handed a pre-configured robotic dog and taught how to control it with a remote. Instead, they will first take it apart. They will examine how each structural plate is connected, whether each servo provides enough torque to support the body weight, and what role every screw plays under mechanical stress. The machine is no longer a black box, but a system waiting to be understood, modified, and redesigned.
我们会让他们拿起锯子，亲手去切割第一块木板。第一次锯的时候会歪，切口会毛糙，尺寸会不精准。然后他们会发现，当结构误差超过两毫米，整个机器狗的腿部运动就会出现偏差。
They will pick up a saw and cut their first wooden board themselves. The first cut will probably be crooked. The edges may be rough, and the dimensions may not be precise. Then they will discover that when a structural error exceeds two millimeters, the entire movement of the robot dog's legs will deviate.
他们会自己打孔，自己量尺寸，自己拧螺丝。排线不会提前贴好，电源模块不会提前焊好。他们会亲手把排线理顺，知道哪一根是信号线，哪一根是供电线。当舵机因为供电不足而抖动时，他们学会测量电压，而不是第一时间怀疑程序。
They will drill holes, measure dimensions, and tighten screws on their own. Wires will not be pre-arranged, and power modules will not be pre-soldered. Students will organize the wiring themselves, learning which wire carries signals and which supplies power. When a servo starts shaking due to insufficient voltage, they will learn to measure the voltage first, rather than immediately blaming the program.
当基础结构完成之后，我们不会停在“能走”这个阶段。真正的升级，才刚刚开始。
Once the basic structure is completed, we will not stop at the stage of "it can walk." The real evolution is only just beginning.
The development of the robotic dog will proceed along two paths.
第一条，是结构的进化。
原始结构可以平地行走，但面对楼梯却显得无力。于是问题出现了：如果想让它爬楼，需要改变什么？腿要更粗吗？关节扭矩要更大吗？重心是不是要下移？电池位置是否需要重新布局？
学生会重新设计腿部结构，增加加强筋，尝试更高强度材料，重新计算受力点。他们会发现，腿部长度增加后，稳定性反而下降，于是必须重新调整步态算法。结构与控制从来不是分开的。
当结构强化到一定程度，他们甚至可以尝试让机器狗具备跳跃能力。那时，他们要考虑的不再只是静态承重，而是瞬间冲击力，是材料疲劳，是动力响应速度。
从木结构到强化结构，从简单行走到爬楼、奔跑、跳跃，这是一次真实的工程升级过程。机器每一次摔倒，都是一次设计修正；每一次站稳，都是一次系统优化。
The first path is the evolution of structure.
The original structure can walk on flat ground, but it struggles when facing stairs. A question naturally arises: if we want it to climb stairs, what needs to change? Should the legs be thicker? Should the joints provide greater torque? Should the center of gravity be lowered? Does the battery placement need to be redesigned?
Students will begin redesigning the leg structure—adding reinforcement ribs, experimenting with stronger materials, and recalculating stress points. They will soon discover that when the legs become longer, stability may actually decrease. As a result, the gait algorithm must also be adjusted. In engineering, structure and control are never separate.
Once the structure becomes strong enough, students may even attempt to give the robotic dog the ability to jump. At that point, they will no longer be dealing only with static load-bearing, but with instantaneous impact forces, material fatigue, and the speed of power response.
From wooden structures to reinforced frameworks, from simple walking to stair climbing, running, and even jumping—this is a real engineering evolution. Every time the machine falls, it becomes a design revision. Every time it stands steadily again, it represents a system optimization.
左滑查看英文版
Swipe left to view the English version
第二条路线，是智能的进化。
当机器能够稳定运动之后，它开始拥有“感知世界”的能力。
学生会为它接入麦克风，让它听见声音；接入语音识别系统，让它理解语言；再接入大模型，让它能够进行自然对话，而不仅仅执行固定指令。
机器狗不再只是被动响应“握手”“坐下”这样的预设动作，而是可以理解更自然的表达，进行互动，甚至根据语境做出不同反应。
随后，他们会加入视觉系统。摄像头接入后，图像识别模型开始运行。机器狗能够识别人脸，能够识别物体，能够跟随目标移动。
语音是输入，视觉是感知，大模型是大脑，电机与结构是身体。当“感知—决策—执行”形成完整闭环时，这台机器开始真正具备系统意义。
The second path is the evolution of intelligence.
Once the machine can move stably, it begins to develop the ability to perceive the world.
Students will connect a microphone so it can hear sounds; integrate a speech recognition system so it can understand language; and then connect it to a large language model so it can engage in natural conversation, rather than simply executing fixed commands.
The robotic dog will no longer passively respond to preset actions like "shake hands" or "sit." Instead, it will be able to understand more natural expressions, interact with people, and even respond differently depending on the context.
Next, students will introduce a vision system. Once a camera is connected, image recognition models begin to run. The robotic dog will be able to recognize faces, identify objects, and follow moving targets.
Voice becomes the input, vision becomes the perception, the large model becomes the brain, and the motors and structure form the body. When perception — decision-making — execution form a complete closed loop, the machine truly begins to function as a system.
左滑查看英文版
Swipe left to view the English version
当结构升级与智能升级在同一系统中融合时，它不再只是课堂实验品，而是一个不断迭代的产品原型。
When structural evolution and intelligent evolution merge within the same system, the machine is no longer just a classroom experiment—it becomes a continuously evolving product prototype.
学生开始思考更真实的问题：
如果把它做成陪伴型机器人，会是什么样？
如果做成巡检机器人，它能进入什么场景？
如果强化结构与智能结合，是否可以成为现实环境中的移动平台？
Students then begin to think about more realistic questions:
What would it look like if it were developed into a companion robot?
If it became an inspection robot, what environments could it operate in?
If stronger structures and intelligent systems are combined, could it become a mobile platform for real-world applications?
Origin Future Engineering Lab
团队由此形成。项目由此诞生。
From here, teams begin to form, and projects emerge.
技术报告、结构图纸、算法逻辑、成本分析逐渐出现。
Technical reports, structural drawings, algorithm logic, and cost analyses gradually take shape.
成熟的成果可以申请专利，可以参加科技竞赛，也可以在条件具备时对接创业资源。
Mature outcomes may lead to patent applications, participation in science and technology competitions, or even connections to entrepreneurial resources when conditions allow.
在过去，做工程是一件门槛极高的事情。设计需要长期积累，算法需要复杂推导，系统调试依赖大量经验。学生往往只能在大学甚至更高阶段，才真正触及“创造”的核心。
In the past, engineering was a field with an extremely high barrier to entry. Design required years of accumulated experience, algorithms required complex derivations, and system debugging depended heavily on deep technical expertise. Students often could only reach the core of real "creation" in university or even later.
但今天，人工智能正在改变这一切。
But today, artificial intelligence is changing this reality.
建模可以借助智能工具快速生成方案，结构可以通过仿真优化，代码可以在大模型辅助下快速迭代，算法逻辑可以通过系统模拟不断验证。工程的复杂度正在被拆解，创新的门槛正在被降低。
Design models can now be generated rapidly with intelligent tools.Structures can be optimized through simulation. Code can be iterated quickly with the assistance of large language models. Algorithmic logic can be repeatedly validated through system simulations. The complexity of engineering is being decomposed, and the barriers to innovation are being lowered.
这意味着，高中阶段已经具备开始“做真实项目”的条件。
This means that high school is already a viable stage to begin building real projects.
学生不再需要等理论完全成熟才动手。他们可以在实践中理解理论，在搭建中理解算法，在调试中理解结构，在失败中理解系统。
Students no longer need to wait until their theoretical knowledge is fully mature before they start building. They can understand theory through practice, understand algorithms through implementation, understand structures through debugging, and understand systems through failure.
当人工智能成为工具，真正重要的能力变成了——提出问题、整合系统、持续迭代。
When artificial intelligence becomes a tool, the truly important abilities become: asking meaningful questions, integrating systems, and iterating continuously.
而这些能力，越早培养越好。
And these abilities are best cultivated as early as possible.
一只机器狗，只是原点。
A robotic dog is only the starting point.
真正被孵化的，是学生面对未来的能力，是跨越工程与智能边界的思维方式，是在人工智能时代主动创造的勇气。
What is truly being incubated is students' capacity to face the future, their ability to think across the boundaries of engineering and intelligence, and their courage to create actively in the age of artificial intelligence.
我们不想培养只会操作工具的人。
We do not want to cultivate people who only know how to operate tools.
我们希望培养的，是能够驾驭工具、重构系统、创造产品的人。
What we hope to nurture are people who can master tools, reconstruct systems, and create products.
而未来，就从这里开始。
And the future begins right here.
第1周｜拆解与认知
目标：
理解机器系统不是黑盒
内容：
• 拆解机器狗结构
• 识别舵机、电源、主控板、信号线
• 分析受力点
• 讲解什么是“系统”
成果：
• 学生画出第一张结构示意图
第2周｜结构基础与材料实验
目标：
理解材料与结构强度
内容：
• 木板切割练习
• 手工打孔、固定结构
• 测试不同厚度木板的承重能力
• 讨论重心与稳定性
成果：
• 制作一个简单承重结构模块
第3周｜舵机与运动控制基础
目标：
理解运动来源
内容：
• 舵机工作原理
• PWM信号基础
• 连接舵机驱动板
• 控制单个关节运动
成果：
• 控制两轴机械臂完成基本动作
第4周｜系统整合测试
目标：
第一次完整组装与调试
内容：
• 结构 + 控制整合
• 电源调试
• 排线整理
• 解决抖动与供电问题
成果：
• 完整运行基础步态
第5周｜爬楼问题分析
目标：
提出升级需求
内容：
• 测试机器狗爬楼失败原因
• 分析扭矩、腿长、重心
• 画升级草图
成果：
• 爬楼结构方案设计图
第6周｜强化腿部结构制作
目标：
结构升级实践
内容：
• 加厚腿部
• 加强筋设计
• 改善连接方式
• 减重优化
成果：
• 强化版本腿部结构
第7周｜步态优化
目标：
理解运动来源
目标：结构与算法配合
内容：
• 调整关节角度
• 测试不同步态
• 稳定性测试
• 改善转弯与重心控制
成果：
• 能稳定爬低台阶
第8周｜结构测试挑战
目标：
实战测试
内容：
• 爬楼挑战
• 连续运行测试
• 记录失败数据
• 改进方案
成果：
• 第一版升级型机器狗
第9周｜语音系统接入
目标：
机器开始“听见”
内容：
• 麦克风模块连接
• 基础语音识别
• 指令触发动作
成果：
• 语音控制坐下/站起
第10周｜大模型接入
目标：
机器开始“理解”
内容：
• 接入大模型API
• 语音转文本
• 文本生成回复
• 语音合成输出
成果：
• 基础对话能力
第11周｜视觉系统接入
目标：
机器开始“看见”
内容：
• 摄像头连接
• 目标识别
• 人脸识别
• 简单目标跟随
成果：
• 跟随指定目标移动
第12周｜智能行为逻辑
目标：
感知—决策—执行闭环
内容：
• 条件触发行为
• 简单决策树
• 多传感器融合
• 行为组合逻辑
成果：
• 可根据环境变化做不同反应
第13周｜产品方向确定
学生分组选择方向：
• 爬楼强化版
• 智能陪伴型
• 巡检机器人
• 机械臂扩展版
输出：
• 产品定位说明书
第14周｜优化与打磨
内容：
• 外观优化
• 结构减重
• 提高稳定性
• 优化交互体验
输出：
• 第二代产品原型
第15周｜技术文档与专利思维
内容：
• 撰写技术说明书
• 绘制结构图
• 成本分析
• 专利基础讲解
输出：
• 项目白皮书
第16周｜路演与展示
内容：
• 项目答辩
• 产品演示
• 讲解技术原理
• 展示未来升级规划
成果：
• 完整产品发布展示
Origin Future Engineering Lab
Phase 1:
Understanding Systems &
Engineering Foundations
(Weeks 1–4)
Week 1 | Disassembly & System Awareness
Objective:
Understand that machine systems are not “black boxes”.
Content:
Disassemble the robot dog structure
Identify servos, power supply, main control board, and signal cables
Analyze load-bearing points
Introduction to the concept of a “system”
Outcome:
Week 2 | Structural Fundamentals & Material Experiments
Objective:
Understand materials and structural strength.
Content:
Wood board cutting practice
Manual drilling and structural fixing
Load-bearing tests with different wood thicknesses
Discussion on center of gravity and stability
Outcome:
Week 3 | Servo Motors & Motion Control Basics
Objective:
Understand the source of mechanical movement.
Content:
Servo motor working principles
Basics of PWM signals
Connecting the servo driver board
Controlling a single joint movement
Outcome:
Week 4 | System Integration Test
Objective:
Complete the first full system assembly and debugging.
Content:
Outcome:
Phase 2:
Structural Upgrades
(Weeks 5–8)
This phase marks the beginning of true engineering evolution.
Week 5 | Stair-Climbing Problem Analysis
Objective:
Identify upgrade requirements.
Content:
Test why the robot dog fails to climb stairs
Analyze torque, leg length, and center of gravity
Sketch upgrade concepts
Outcome:
Week 6 | Reinforced Leg Structure Construction
Objective:
Implement structural upgrades.
Content:
Outcome:
Week 7 | Gait Optimization
Objective:
Integrate structure with control algorithms.
Content:
Outcome:
Week 8 | Structural Challenge Test
Objective:
Conduct real-world testing.
Content:
Outcome:
Phase 3:
Intelligent Upgrades
(Weeks 9–12)
This stage introduces AI system capabilities.
Week 9 | Voice System Integration
Objective:
Enable the robot to hear.
Content:
Outcome:
Week 10 | Large Model Integration
Objective:
Enable the robot to understand.
Content:
Outcome:
Week 11 | Vision System Integration
Objective:
Enable the robot to see.
Content:
Camera integration
Object recognition
Face recognition
Simple target tracking
Outcome:
Week 12 | Intelligent Behavior Logic
Objective:
Build a Perception → Decision → Execution loop.
Content:
Outcome:
Phase 4:
Productization & Entrepreneurial Thinking
(Weeks 13–16)
Transitioning from project to product.
Week 13 | Product Direction Selection
Students choose one direction in teams:
Stair-Climbing Enhanced Version
Intelligent Companion Robot
Inspection Robot
Robotic Arm Expansion Version
Output:
Week 14 | Optimization & Refinement
Content:
Output:
Week 15 | Technical Documentation & Patent Thinking
Content:
Writing technical documentation
Drawing structural diagrams
Cost analysis
Introduction to patent fundamentals
Output:
Week 16 | Final Pitch & Showcase
Content:
Outcome:
左滑查看英文版
Swipe left for the English version.
最终能力提升
Final Competency Development
16周结束后，学生将具备：
材料与结构工程理解
电控系统整合能力
AI语音与视觉接入能力
系统整合思维
产品化思维
创业表达能力
After completing the 16-week program, students will have developed:
An understanding of materials and structural engineering
The ability to integrate electronic control systems
Skills in AI voice and vision integration
Systems integration thinking
Product development mindset
Entrepreneurial communication and presentation skills
撰文：Sam老师
拍摄：Marty老师
排版编辑：新媒体部
繁体
