SBIR Phase I: An Online Reliability and Risk Analysis Software Product

Gensym Corporation的这个小型企业创新研究第一阶段项目将调查开发在线、实时可靠性和风险分析软件的可行性，该软件将分析过程数据以确定设备和系统功能的可靠性，评估设施中的当前风险，报告并可能警告操作员和管理人员系统功能和设备的风险，并建议纠正措施以保持对风险的控制。对于许多组织来说，在风险缓解和安全方面的支出是巨大的。通过更仔细地监控风险并在最糟糕的情况发生之前采取纠正措施，可以降低风险缓解成本。监控能力越强，风险发现和规避行动之间的反馈环路越紧密，节省的资金就越多。这项研究将证明，贝叶斯网络为大多数可靠性和风险分析模型提供了核心表示，可以利用系统的可分解性来提供实时评估，并且可以利用学习方法来估计和重新校准分析。在线风险管理能力在许多行业具有很高的价值：环境监测、有毒废物处理、清理和运输、核能发电、近海石油生产和运输、紧急响应和紧急疏散、关键电信、交通运输、灵活制造、证券交易、信用欺诈预防等许多行业。

This Small Business Innovation Research Phase I project from Gensym Corporation will investigate the feasibility of developing online, real time reliability and risk analysis software that will analyze process data to ascertain the reliability of equipment and system functions, assess the current risk in the facility, report and possibly alarm operators and managers of the risk to system functions and equipment, and recommend corrective actions to maintain control over risk. For many organization, expenditure on risk mitigation and safety is significant. Risk mitigation costs can be reduced by more carefully monitoring risk and taking corrective action before worst case scenarios develop. The greater the ability to monitor and the tighter the feedback loop between risk discovery and evasive action, the more significant the savings. This research will demonstrate that Bayesian networks provide a core representation for most reliability and risk analysis models, that system decomposability can be exploited to provide real time assessment, and that learning methods can be utilized to estimate and recalibrate the analysis. The value for an online risk management capability is high in many industries: Environmental monitoring, toxic waste processing, cleanup and shipment, nuclear power generation, offshore oil production and shipment, emergency response and emergency evacuation, critical telecommunications, transportation, flexible manufacturing, securities trading, credit fraud prevention, and many others.