Kitchen Impossible Mission
厨房不可能任务
A ramen-making training simulator built with AI vibe coding, based on real kitchen experience at a Japanese restaurant in the Netherlands.
基于荷兰日本拉面馆真实厨房经验，用AI vibe coding制作的拉面制作培训模拟器。
The Story ↓
I worked part-time at a Japanese ramen restaurant in the Netherlands, starting as a server and eventually moving into the kitchen.
The restaurant had high staff turnover and labor shortages, so the owner had already standardized the cooking process to minimize training costs. The cooking itself wasn't hard — the real challenge was doing it fast and error-free.
During peak hours, 8 bowls of ramen might be waiting while you can only handle 3 at once. Each bowl has different flavors, toppings, and noodle types. One mistake means the whole bowl gets thrown out.
I realized the biggest challenge in the kitchen isn't cooking — it's managing multiple tasks simultaneously under time pressure while maintaining accuracy. So I turned this experience into a training game.
去年我在荷兰的一家日本拉面馆做兼职，从最开始在前厅做服务员，到后来尝试在厨房做饭。
拉面馆人员流动大、劳动力短缺，老板为了减少培训与雇佣成本，早已将制作流程简化标准化。做饭本身不难，难的是怎么做得又快又不出错。
高峰时段同时有8碗拉面等待制作，但你最多只能兼顾3碗。每碗拉面的口味、配料和面条种类都不一样。一旦犯错，整碗面都要扔掉重做。
我逐渐发现，厨房里最大的挑战并不是做饭本身，而是在有限时间内同时管理多个任务，并保持准确率。于是我把这段经历做成了一个培训游戏。
The Game ↓
The game recreates the real cooking process of the restaurant's signature vegan miso ramen, with the interface designed to match the actual kitchen layout.
Players follow 8 steps from ingredient preparation to final plating: frying seitan, heating broth, boiling noodles, blanching vegetables, draining, assembling the bowl, stacking 10+ toppings, and wiping the rim for service.
The industrial standard is to complete plating and serve within 1 minute. Through repeated practice, players build muscle memory for the workflow and kitchen flow, so they can stay calm during actual shifts.
Currently only one ramen flavor is implemented — a demo. The real challenge at work is memorizing and switching between 10 different recipes on the fly.
游戏还原了餐厅招牌素食味噌拉面的真实制作流程，界面按照餐厅实际的厨房布局设计。
玩家需要按照8个步骤完成制作：素肉炸制、高汤唤醒、面条制备、鲜蔬焯水、面水控干、暖碗盛装、美学封顶、擦边出餐。
实际的工业标准是从摆盘到送到顾客桌前在1分钟内完成。通过反复练习，玩家可以熟悉制作流程和厨房动线，训练肌肉记忆，在实际工作中不慌乱。
目前仅上线了一款拉面，更像一个Demo。实际工作中最大的挑战是要记住并随时切换10种不同口味拉面的制作流程和配料组合。
Process ↓
The game was built entirely through vibe coding — using natural language to describe what I wanted, and letting AI generate the code.
I started with hand-drawn sketches to define the core features and layout, then fed the sketches along with text descriptions to Google AI Studio, which generated the first working UI.
After two rounds of iteration — tightening the ingredient layout, clarifying functional zones, and improving interaction feedback — the game reached its current form.
Having real domain knowledge was more important than coding skills. Every step, every topping position comes from actual kitchen experience. AI writes the code, but domain expertise is the real moat.
游戏完全通过vibe coding构建——用自然语言描述需求，让AI生成代码。
我先用手绘草图确定核心功能与布局，然后将草图和文字描述提供给Google AI Studio，AI直接生成了第一版可用的UI。
经过两轮迭代——食材布局更紧凑、功能分区更清晰、操作反馈更直观——游戏达到了目前的效果。
懂领域比懂技术更关键。每一步操作、每个配料位置都来自真实经验。AI帮你写代码，领域知识才是你独有的护城河。
Hand-drawn sketch / 手绘草图
First AI-generated UI / AI生成初版