ITR - (ASE + NHS) - (int): Intelligent Human-Machine Interface & Control for Highly Automated Chemical Screening Processes

危险化学制剂对人体细胞和细胞功能影响的高通量毒性筛选（检测）是国际上迅速发展的生物技术产业。先进的机器人已经取代了人工操作员和人工控制的筛选过程，以促进对化学品和毒素的安全、快速和准确的评估。在筛选过程中，人类操作员的角色已经转变为多个机器人在多条工艺线上同时进行实验的监督控制器。操作员的任务现在包括监控机器人状态、检测错误和干预故障模式恢复的过程控制。操作人员在高工作量和时间压力下正确排序实验，并在实验没有按计划进行时迅速做出改变。有效执行任务所需的信息急剧增加；任务负荷现在主要是认知和身体；在危险的化学操作中，有必要实现高水平的态势感知。所有这些变化都对操作人员的压力和工作健康产生了重大影响。在这个项目中，PI和他的团队将开发一个智能/自适应的人机界面，以支持筛选过程监督员在安全有效、高时间压力和高风险的分布式控制、自动化化学和毒性测试中的新角色。该技术的发展将基于对实际化学筛选过程中监督控制器行为的认知建模，以及在（模型）设计阶段和化学过程运行期间使用不同交互信息显示设计方案对操作员性能的模型预测。PI将为操作人员提供原型控制接口和共享态势感知显示，集成并显示适应操作人员并发性能需求和功能（生理）状态的过程输出数据。这项工作还将涉及为在不同网络通信条件下远程操作自动筛选过程制定协议，以便为“初创”公司和有关键筛选需求（例如反恐工作）的发展中国家提供访问。此外，智能界面内容和远程过程控制场景将由自动化机器人（机械）“健康”监测系统提供信息，并根据其进行调整。该项目的成果将包括在远程过程控制系统中体现的新型原型智能，自适应接口技术。还将开发先进的计算工具，根据生理数据实时分类操作员的功能状态，并将这些信息与筛选任务的工作量和性能联系起来。生理和性能数据模型将作为构建认知模型的基础，以促进高度准确的操作员性能预测，并为远程筛选过程控制提供有效的动态接口配置。更广泛的影响：本研究将影响并提高高通量化学剂筛选的安全性和有效性，这得益于复杂自动化系统分布式网络控制的新方法，以用户为中心的自适应/智能接口设计，以及通过基于操作员功能状态的过程信息集成和管理来支持复杂系统操作员的态势感知和决策。这项工作将增加新公司和发展中国家获得高度专业化和昂贵的自动化化学筛选技术的机会，可能加速新生物技术的发展。此外，该研究将为参与该项目的研究生提供专门培训，并通过教师开发与该项目相关的新课程模块，将其整合到现有的计算机科学，电气工程和工业工程课程中。

High-throughput toxicity screening (testing) of dangerous chemical agents for effects on human cells and cell functions is a rapidly developing international, biotechnology industry. Advanced robots have replaced human operators and manual control of screening processes to promote safe, quick and accurate assessment of chemicals and toxins. In the screening process, the role of the human operator has changed to that of supervisory controller of multiple robots manning multiple process lines and performing simultaneous experiments. The operator's task now includes monitoring robot states, detecting errors, and intervening in process control for failure mode recovery. Operators are under high workload and time stress to properly sequence experiments and to quickly make changes if they do not progress as planned. The information requirements for effective performance have expanded dramatically; task workload is now primarily cognitive vs. physical; and there is a need to achieve high levels of situation awareness in dangerous chemical operations. All of these changes have had a major affect on operator stress and work health.In this project, the PI and his team will develop an intelligent/adaptive, human-machine interface to support the new role of screening process supervisors in safe and effective, distributed control of high time stress and high risk, automated chemical and toxicity testing. Development of this technology will be based on cognitive modeling of supervisory controller behaviors during actual chemical screening processes and model predictions of operator performance with different interactive information display design alternatives during the (model) design phase and during chemical process run-time. The PI will prototype control interfaces and shared situation awareness displays for operators that integrate and display process output data adapted to operator concurrent performance needs and functional (physiological) states. The work will also involve creating protocols for long-distance, remote operation of automated screening processes under varying network communication conditions to provide access to "start-up" companies and developing nations with critical screening needs (e.g., anti-terrorism work). In addition, the intelligent interface content and the remote process control scenario will be informed by, and adapted based on, an automated robot (mechanical) "health" monitoring system. Outcomes of the project will include a prototype novel intelligent, adaptive interface technology as embodied in a remote process control system. Advanced computational tools will also be developed to classify operator functional states in real-time, based on physiological data, and to relate this information to screening task workload and performance. Physiological and performance data models will be used as a basis for structuring the cognitive model to promote highly accurate operator performance predictions and facilitate effective dynamic interface configuration for remote screening process control.Broader Impacts: This research will impact and enhance the safety and effectiveness of high-throughput, chemical agent screening, thanks to a combination of new approaches to distributed network control of complex automated systems, user-centered design of adaptive/intelligent interfaces, and support of complex system operator situation awareness and decision making through process information integration and management based on operator functional states. The work will result in increased access for new companies and developing countries to highly specialized and expensive automated, chemical screening technologies potentially accelerating the development of new biotechnologies. In addition, the research will provide specialized training for graduate students participating in the project and through faculty development of new course modules, related to the project, integrated in existing computer science, electrical engineering, and industrial engineering curriculums.