访校团探路者课程｜原点未来工程研究所：从0到1构建机器人系统，GT1号与机器人工程师的诞生
←左滑查看英文版
←Swipe left to view the English version.
在高藤「探路者课程」框架下，原点未来工程研究所作为以工程实践与创业精神为核心的科研实验室，学生以“科研工程师”身份参与真实工程项目——从需求定义、方案设计，到机械结构实现、软件开发与系统调试，完整经历一套机器人研发流程。
而GT1号，正是这一体系下诞生的第一个工程化成果。
在短短5周内，研究团队基于废弃材料完成了一只具备基础机械结构与初级智能能力的“通用型机械狗”原型。
这一次，我们走进原点未来工程研究所，与 Sam 老师及学生团队进行了一次更深入的对话，试图还原这段快速生长背后的真实路径。
在后续规划中，GT系列将向更完整的“通用型机器人系统”迭代，包括自主导航、身份识别与复杂任务执行（如实际场景服务应用）。
在原点未来工程研究所中，机器人原型从零开始构建。
材料方面，研究团队选择了校园内的废弃装修木板作为GT1号的原型载体，在没有工业化外壳与标准化组件的前提下，结构设计反而获得了更大的开放空间。
在缺乏标准化零部件的条件下，团队需要自行完成切割、打孔、测量与结构拼接，构建完整的机械结构系统。
这不仅是动手能力的训练，更是一种工程决策能力的建立——如何在有限条件下完成选材、连接与结构构建，使系统具备基本的可运行性。
（高藤校园内闲置的木板）
0成本、易裁切、规格丰富、适合快速试错。
在这个阶段，
他们正在经历的是
实验室中的原型开发，
而非课堂中的制作练习。
在原点未来工程研究所的框架中，机器人并不只是一个单点技术产品，而是一个由多学科协同支撑的复杂系统，其背后涉及材料科学与机械结构等多个上下游机器人产业链工程环节。
从“被搭建”到“稳定行走”，过程中GT1号多次出现经历结构失效问题，包括受力不均、重心偏移与运动抖动。
通过反复排查，研究团队逐步将问题定位于结构设计本身，并据此展开新一轮迭代。
重新设计腿部结构，扩大支撑范围
为脚部增加防滑垫，提升接触稳定性
持续优化连接方式，寻找更合理的受力路径
在现场，为了解决重心不稳的问题，
同学们再次尝试通过优化腿部结构进行调整。
在这个过程中， 研究成员们不仅完成了结构优化，也逐步强化了对材料与结构工程的理解：
如果说机械结构是“身体”，
那么软件与AI系统则是“神经”。
在GT1号的开发中，他们围绕软件、人工智能与网络通信等核心技术，构建起GT1号的智能系统，使其能够在控制、电力与机械执行之间实现协同，并具备基础人机交互能力。
随着功能叠加，问题也逐渐出现：指令失效、动作错乱、模块冲突等问题开始集中出现。
研究团队于是开始自主解决问题：
通过AI辅助定位问题，并从代码逻辑、通信接口到执行链路逐层排查优化；
机械与硬件团队同步调整结构与电路，实现软硬件协同修正。
在现场，我们也看到了熟悉的身影——Ian。他正与同伴一起进行 Debug。
这一过程推动项目
从功能开发走向系统整合，
也让研究成员逐步建立起
跨模块工程思维。
5周完成一只机械狗原型，并不是终点。
在原点未来工程研究所，GT1号的开发过程，是一条完整的机器人产业链。团队成员在项目中接触到的，不只是单一技能，而是从结构设计、硬件实现到软件系统与人工智能的跨领域整合能力。
这也正是原点未来工程研究所
设立的核心目标——
在持续推进GT1号的过程中，学生的能力也在同步生长：
从最初对材料与结构的理解，到逐步掌握电控系统的搭建与整合；
从接入语音与视觉等AI能力，到能够从系统层面理解不同模块之间的协同关系，并形成完整的系统整合思维。
同时，在不断迭代与对外展示的过程中，他们也开始具备产品化思考与基础的技术表达能力。
在下一阶段，GT1号还将继续迭代：机械臂、更高负载能力、更丰富的交互方式（如表情显示、人脸识别）等功能都已在规划之中。
未来，他们或将带着GT1号走向竞赛舞台，也可能在更长远的路径中，成为工程师、开发者，甚至进入更广阔的科技领域。
而这一切的起点，是在原点未来工程研究所，从零开始构建一套真实运行的机器人系统。
写在最后
从一块废弃木板，到一只具备初步智能能力的机械系统，GT1号的诞生，并不仅仅是一次项目成果的呈现。它更像是一种“研究范式”的验证——
在原点未来工程研究所，
学生以科研人员的身份
进入真实工程语境，
在不确定中建构系统，
在实践中形成认知。
在AI与工程深度融合的时代，“未来”不再只是被讲授的知识,而是可以被一群年轻的工程师，从零开始，亲手构建出来。
Under the framework of SMCS"Pathfinder Program," the Origin Future Engineering Lab serves as a research laboratory centered on engineering practice and entrepreneurial spirit. Students participate in real engineering projects as "research engineers"—from defining requirements and designing solutions, to implementing mechanical structures, developing software, and debugging systems—experiencing the complete robotics development process.
GT1 is the first engineered outcome born from this system.
In just five weeks, the research team completed a "general-purpose robotic dog" prototype with basic mechanical structures and preliminary intelligent capabilities, built from discarded materials.
This time, we visited the Origin Future Engineering Lab for an in-depth conversation with Teacher Sam and the student team, aiming to uncover the real path behind this rapid growth.
Moving forward, the GT series is planned to evolve into a more comprehensive "general-purpose robotic system," featuring autonomous navigation, identity recognition, and the ability to perform complex tasks (e.g., real-world service applications).
At the Origin Future Engineering Lab, robotic prototypes are built from the ground up.
In terms of materials, the research team chose discarded wooden boards from the campus as the prototype carrier for GT1. Without an industrial shell or standardized components, the structural design gained greater freedom and openness.
Lacking standardized parts, the team had to handle cutting, drilling, measuring, and structural assembly themselves, building a complete mechanical system from scratch.
This is not just hands-on skill training; it is also the development of engineering decision-making ability—how to select materials, make connections, and build structures under constraints so that the system achieves basic operability.
(discarded wooden boards on the SMCS campus)
zero cost, easy to cut, available in various sizes, and perfect for rapid prototyping.
At this stage,
what they are experiencing is
prototype development in a lab setting,
not a hands-on exercise in a classroom.
Within the framework of the Origin Future Engineering Lab, a robot is not merely a single-point technical product, but a complex system supported by multidisciplinary collaboration. Behind it lie multiple upstream and downstream robotics engineering processes, including materials science and mechanical structures.
From being "assembled" to "walking stably," GT1 experienced multiple structural failures along the way, including uneven force distribution, center-of-gravity shifts, and motion-induced shaking.
Through repeated investigation, the research team gradually pinpointed the root cause in the structural design itself and accordingly launched a new round of iteration:
Redesigning leg structures to expand the support area
Adding anti-slip pads to the feet to improve contact stability
Continuously optimizing connection methods to find better load paths
On site, to address the issue of an unstable center of gravity, the students once again attempted to make adjustments by optimizing the leg structure.
Throughout this process, the team members not only completed structural optimization but also gradually deepened their understanding of materials and structural engineering:
If the mechanical structure is the "body,"
then the software and AI system
are the "nervous system."
In the development of GT1, the team built its intelligent system around core technologies such as software, artificial intelligence, and network communication, enabling GT1 to achieve coordination between control, power, and mechanical execution, as well as basic human‑robot interaction capabilities.
As more functions were added, problems gradually emerged: command failures, incorrect actions, module conflicts, and other issues began to appear intensively.
The research team then started solving problems on their own, using AI assistance to locate issues, and systematically debugging and optimizing from code logic, communication interfaces, down to the execution chain.
Meanwhile, the mechanical and hardware teams adjusted the structure and circuits in parallel, achieving coordinated software‑hardware fixes.
On site, we also saw a familiar face—Ian. He was debugging with a teammate.
This process pushed the project
from functional development
toward system integration,
and gradually helped the team members
build cross‑module engineering thinking.
Completing a robotic dog prototype in five weeks is not the finish line.
At the Origin Future Engineering Lab, the development process of GT1 represents a complete robotics industry chain. What the team members encounter in the project is not just a single skill, but the cross‑disciplinary integration of structural design, hardware implementation, software systems, and artificial intelligence.
This is precisely the core goal of the lab—
As GT1 continues to evolve, the students' abilities grow in parallel:
from their initial understanding of materials and structures, to gradually mastering the construction and integration of electronic control systems;
from incorporating AI capabilities such as voice and vision, to understanding at a system level the collaborative relationships between different modules and developing complete system integration thinking.
At the same time, through continuous iteration and external presentations, they also begin to develop product‑oriented thinking and basic technical communication skills.
In the next phase, GT1 will continue to evolve: a robotic arm, higher load capacity, richer interaction methods (like facial expression display, face recognition) are already in the pipeline.
In the future, they may take GT1 to competitions, or on a longer path, become engineers, developers, or even enter broader technology fields.
And the starting point of all this is at the Origin Future Engineering Lab—building a real, functioning robotic system from scratch.
Epilogue
From a discarded wooden board to a mechanical system with preliminary intelligence, the birth of GT1 is not merely the presentation of a project outcome. It is more like a validation of a "research paradigm"—
At the Origin Future Engineering Lab,
students step into real engineering contexts as researchers,
building systems amidst uncertainty,
and forming understanding through practice.
In an era where AI and engineering are deeply integrated, the "future" is no longer just knowledge to be taught—but something that can be built from scratch, with their own hands, by a group of young engineers.
声明： 本文内容为国际教育号作者发布，不代表国际教育网的观点和立场，本平台仅提供信息存储服务。
最新评论