Collaborative Research: DASS: Accountable Software Systems for Safety-Critical Applications

合作研究：DASS：用于安全关键应用的负责任的软件系统

• 批准号: 2131511
• 负责人: Lu Feng
• 金额: $ 37.49万
• 依托单位: University of Virginia Main Campus
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2021
• 资助国家: 美国
• 起止时间: 2021-10-01 至 2024-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2131511&HistoricalAwards=false
• 关键词: Collaborative; Research; DASS; Accountable; Software

Safety-critical software systems are entering the market in large numbers and are expected to transform many industries including healthcare, transportation, manufacturing, and others. In response to the rising societal impact and complexity of software systems, lawmakers and regulatory authorities are implementing new laws, regulations, and guidelines to hold software accountable for its harmful effects. These legal approaches differ across jurisdictions and across application domains, and they will evolve over time as lawmakers and regulators continue to study and address emerging software capabilities. Despite this mounting regulatory pressure, state-of-the-art software-design methodologies are deficient at providing the desired accountability in safety-critical systems. The project’s novelties are twofold: (1) developing principled approaches and tools for assuring and demonstrating accountability of safety-critical software systems with respect to laws and regulations that evolve over time, and (2) advancing a legal framework that harmonizes regulatory oversight of software systems across heterogeneous safety-critical domains. The project’s impacts are facilitating the design of safety-critical software systems that are accountable with respect to various regulations, and providing legal insight on how to extend or amend current regulatory approaches to enhance software accountability. In addition, the investigators will organize a series of interdisciplinary workshops and symposiums to bring together experts in software design and law to discuss open research questions and potential solutions to software accountability. The investigators also plan to develop new course materials in computer science and law to integrate the proposed research outcomes, and actively recruit underrepresented students for positions in the proposed project.The project includes three research thrusts that seek to make fundamental contributions to both software design and law. The first thrust creates novel approaches and tools for developing compositional dynamic assurance cases throughout the software development lifecycle to assure and demonstrate accountability. The second thrust develops novel formal-verification techniques for generating provable and certifiable regulation compliance guarantees, which can be used as evidence in assurance cases. The third thrust develops legal insight on how lawmakers and regulators should extend or amend current regulatory approaches to incorporate advances in software accountability methods.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

安全关键型软件系统正在大量进入市场，并有望改变许多行业，包括医疗保健、运输、制造业和其他行业。为了应对软件系统日益增长的社会影响和复杂性，立法者和管理当局正在实施新的法律、法规和指导方针，以使软件对其有害影响负责。这些法律方法在不同的司法管辖区和不同的应用领域是不同的，随着立法者和监管者继续研究和解决新兴的软件功能，它们将随着时间的推移而发展。尽管监管压力越来越大，但最先进的软件设计方法在提供安全关键系统所需的问责制方面仍存在不足。该项目的新颖之处有两个：(1)开发原则性的方法和工具，以确保和演示安全关键软件系统的责任，这些责任与随着时间的推移而发展的法律和法规有关，以及(2)推进一个法律框架，以协调跨异构安全关键领域的软件系统的监管监督。该项目的影响是促进对各种法规负责的安全关键软件系统的设计，并提供关于如何扩展或修改当前监管方法以增强软件责任的法律见解。此外，调查人员将组织一系列跨学科研讨会和专题讨论会，汇集软件设计和法律方面的专家，讨论开放式研究问题和软件问责制的潜在解决方案。研究人员还计划开发新的计算机科学和法律课程材料，以整合拟议的研究成果，并积极招募代表性不足的学生担任拟议项目的职位。该项目包括三个研究重点，旨在为软件设计和法律做出根本性的贡献。第一个推力创建了新的方法和工具，用于在整个软件开发生命周期中开发组合动态保证用例，以确保并演示责任。第二个重点是开发新的形式验证技术，用于生成可证明和可证明的法规遵从性保证，这些保证可以用作保证案例中的证据。第三个要点发展了关于立法者和监管者应如何扩展或修改当前监管方法以纳入软件问责方法的进步的法律见解。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

项目成果

Towards Developing Safety Assurance Cases for Learning-Enabled Medical Cyber-Physical Systems

为学习型医疗网络物理系统开发安全保证案例

• 发表时间: 2023
• 期刊: Proceedings of the Workshop on Artificial Intelligence Safety 2023 (SafeAI 2023
• 作者: Maryam Bagheri;Josephine Lamp;Xugui Zhou;Lu Feng;Homa Alemzadeh
• 通讯作者: Homa Alemzadeh