Since the inception of human research studies, researchers often need to interact with participants on a set schedule to collect data. While some human research is automated, most is not; which costs researchers both time and money. Usually, user-provided data collection consists of surveys administered via telephone or email. While these methods are simplest, they are tedious for the survey administrators, which could incur fatigue and potentially lead to collection mistakes. A solution to this was the creation of "chatbots". Early developments relied on mostly rule-based tactics (e.g. ELIZA), which were suitable for uniform input. However, as the complexity of interactions increases, rule-based systems begin breaking down since there exist a variety of ways for a user to express the same intention. This is especially true when tracking states within a research study (or protocol). Recently, natural language processing (NLP) models and, subsequently, virtual assistants have become increasingly more sophisticated when communicating with users. Examples of these efforts range from research studies to commercial health products. This project leverages recent advancements in conversational artificial intelligence (AI), speech-to-text, natural language understanding (NLU), and finite-state machines to automate protocols, specifically in research settings. This application must be generalized, fully customizable, and irrespective of any research study. These parameters allow new research protocols to be created quickly once envisioned. With this in mind, I present SmartState, a fully-customizable, state-driven protocol manager combined with supporting AI components to autonomously manage user data and intelligently determine the intention of users through chat and end device interactions to drive protocols.

翻译：自人类研究诞生以来，研究者通常需要按预定时间表与参与者交互以收集数据。尽管部分人类研究已实现自动化，但多数尚未完成；这既耗费研究者时间又增加经济成本。通常，用户提供的数据收集通过电话或电子邮件进行的问卷实现。虽然这些方法最为简便，但对问卷管理者而言非常繁琐，可能导致疲劳并引发收集错误。对此的解决方案是创建"聊天机器人"。早期发展主要依赖基于规则的策略（如ELIZA），适用于统一输入的场景。然而随着交互复杂度的提升，规则系统开始失效——因为用户表达相同意图的方式存在多样性。这在追踪研究（或协议）中的状态时尤为突出。近年来，自然语言处理（NLP）模型及随后的虚拟助手在与用户交流方面日趋成熟。相关应用涵盖从学术研究到商业健康产品的广泛领域。本项目利用对话式人工智能（AI）、语音转文本、自然语言理解（NLU）和有限状态机的最新进展，对研究场景中的协议进行自动化处理。该应用必须具有泛化性、完全可定制性，且不局限于任何特定研究。这些特性使得新研究协议可在构想后快速创建。基于此，我提出了SmartState——一种完全可定制的状态驱动型协议管理器，结合辅助AI组件，通过聊天和终端设备交互自主管理用户数据并智能识别用户意图，从而驱动协议执行。