Speaker
Description
The first phase of the High Energy Photon Source (HEPS) was completed in 2025. HEPS now offers 14 public beamlines and one optics test beamline, covering a broad range of research areas. HEPS calls for experimental proposals from researchers worldwide. The proposal management process encompasses a comprehensive sequence of stages: submission, multi‑tiered review, beamtime approval, scheduling, on‑site execution, and outcome reporting. To address these needs, we have developed a scientific user service software that enables end‑to‑end management of experimental proposals.
Starting in September 2025, HEPS conducted three rounds of pilot proposal calls via the User Service System, resulting in over 200 approved pilot proposals, more than 4,700 hours of scheduled beamtime, and over 700 user visits. Building on this experience, the user service system was further refined, and the first call for general proposals was launched in March 2026.
Given the rapid advancements in artificial intelligence (AI), we have further designed a human‑computer collaborative platform powered by large language models. This platform seamlessly integrates AI capabilities throughout the entire workflow—from proposal drafting and review to evaluation, experiment planning, and knowledge consolidation. By incorporating intelligent assistance, the system will significantly enhance both the efficiency and quality of proposal processing.
This work outlines the requirements analysis for the user service software, describes the developed framework, elaborates on the core functional modules, and presents specific design strategies for AI‑enhanced implementation.
