Alibaba's Qianwen helps National Astronomical Observatory build telescope simulation system
阿里：千问办公协助国家天文台科研团队，耗时仅 3 天打造科研级望远镜仿真系统、成本不足千元
For users in the scientific research field, this demonstrates how an AI assistant can significantly reduce the cost and time required to build complex simulation systems, making advanced research tools more accessible.
Alibaba's Qianwen Office assisted the National Astronomical Observatory's research team in creating a digital simulation system for a large-aperture scientific telescope. The system was built in just 3 days at a cost of less than US$1,000. Previously, similar systems took about three months and cost tens of thousands of US dollars. The system integrates telescope components, sensor states, and observation environment information into a standard MCP interface for agent monitoring and calls. Agents can autonomously complete a closed loop of 'state perception, task planning, solution generation, and process verification' by combining scientific target priorities, telescope real-time status, and target observable windows. The agent framework has already been integrated with 'SiTian' Pathfinder and 'SiTian' prototype machines, and has successfully warned of 8 early supernova candidates, triggering follow-up observations for 2 of them under suitable weather conditions.
阿里：千问办公协助国家天文台科研团队，耗时仅 3 天打造科研级望远镜仿真系统、成本不足千元
IT之家 9 月 18 日消息，阿里巴巴今日宣布，近日，千问办公协助国家天文台科研团队打造大口径科研级望远镜数字化仿真系统。 耗时仅 3 天，成本不足千元 。 而在过去，同类系统需委托软件供应商开发，周期约三个月，成本高达数万元， 千问办公将这一搭建门槛降低了两个数量级 。 科研级望远镜资源有限，真实设备难以承担反复试错的代价。在仿真系统中提前排演，是减少真实设备试错代价的可行路径之一。 官方表示，千问办公协助国家天文台相关科研团队搭建大口径科研级望远镜数字化仿真系统，并将望远镜各组件、传感器状态及观测环境信息 整合为标准 MCP 接口，供 Agent 监测与调用 。相当于把望远镜组件等变成了一排“通用插座”，Agent 即插即用，既看得见状态，也调得动设备。 ▲ 千问办公演示页面 最终，Agent 可以进一步综合科学目标优先级、望远镜实时状态以及目标可观测窗口，自主完成“状态感知 —— 任务规划 —— 方案生成 —— 流程验证”的闭环。 IT之家获悉，不止于仿真系统中的模拟，该 Agent 框架已接入“司天”探路者和“司天”原型机。截至目前，相关智能体已经预警 8 颗极早期超新星候选，其中 2 颗在天气条件允许的情况下触发了后续观测。
