Verification of Autonomous Systems
暫譯: 自主系統的驗證

Redfield, Signe, Sofge, Donald, Seto, Mae

Verification of Autonomous Systems
  • 出版商: Springer
  • 出版日期: 2026-04-11
  • 售價: $7,060
  • 貴賓價: 9.5 折 $6,707
  • 語言: 英文
  • 頁數: 353
  • 裝訂: Hardcover - also called cloth, retail trade, or trade
  • ISBN: 3031885457
  • ISBN-13: 9783031885457
  • 相關分類: 自駕車

    • 海外代購書籍(需單獨結帳)

商品描述

This book is a comprehensive guide to current practical and theoretical understanding of verification of autonomous systems, helping users find the tools and techniques they need to address this challenging problem. Autonomous systems are transitioning out of the lab and into the commercial and industrial space in ever increasing numbers. Verification provides the assurance case necessary for deployment of autonomous systems, but we lack guidance, certifications, and standards to help a business determine whether their autonomous system is safe, secure, or reliable. A community has formed to develop tools, techniques, and processes to solve this problem, but the existing literature largely focuses on autonomy design tools and techniques rather than on tools and techniques that can be used to verify an existing autonomy design. This book fills that gap.

It directly connects specific challenges faced by verification agents and autonomy developers with the research and topics most relevant to their concerns:

    Specific challenges faced by stakeholders and researchers are cross-referenced to chapters addressing the relevant tools and techniques. Individual aspects of the verification process like requirements and specification development, model tests, and integration tests are mapped into technical challenges related to modeling, testing, abstraction and verification tools. Each chapter is tied to both the stakeholder challenges and the technical challenges it addresses.
Verification processes, industrial needs, and issues with verification of learning systems are addressed across all domains, encompassing platforms from aircraft and spacecraft to marine systems and ground vehicles operating in industrial, military, commercial and household applications. Broad in scope, this reference ties advances in formal analysis, hardware testing, verification process modifications, and design and evaluation tools to the needs of verification researchers and professionals.

商品描述(中文翻譯)

本書是一本全面的指南，針對當前自主系統驗證的實務與理論理解，幫助使用者找到解決這一挑戰性問題所需的工具和技術。自主系統正以不斷增加的數量從實驗室轉向商業和工業領域。驗證提供了部署自主系統所需的保證案例，但我們缺乏指導、認證和標準來幫助企業判斷其自主系統是否安全、可靠或可信。為了解決這個問題，一個社群已經形成，致力於開發工具、技術和流程，但現有文獻主要集中在自主設計工具和技術上，而不是用於驗證現有自主設計的工具和技術。本書填補了這一空白。

本書直接將驗證代理和自主開發者面臨的具體挑戰與最相關的研究和主題聯繫起來：

- 利益相關者和研究人員面臨的具體挑戰與針對相關工具和技術的章節進行交叉參考。
- 驗證過程中的個別方面，如需求和規範開發、模型測試和整合測試，映射到與建模、測試、抽象和驗證工具相關的技術挑戰。
- 每一章節都與利益相關者的挑戰和其所解決的技術挑戰相連結。

驗證過程、工業需求以及學習系統驗證的問題在所有領域中都有所涉及，涵蓋從飛機和太空船到海洋系統和地面車輛的各種平台，這些平台在工業、軍事、商業和家庭應用中運行。本參考書範圍廣泛，將形式分析、硬體測試、驗證過程修改以及設計和評估工具的進展與驗證研究人員和專業人士的需求相結合。

作者簡介

Signe Redfield is the Director of the Laboratory for Autonomous Systems Research at the Naval Research Laboratory (NRL). She has served as Co-Chair for the IEEE Technical Committees for Verification of Autonomous Systems (TC-VAS) and Performance Evaluation and Benchmarking of Robotic and Automation Systems (TC-PEBRAS). She co-authored the first IEEE RAS standard, 1872-2015, and served as Secretary for standard 1872.1-2024, "Robot Task Representation". She was detailed to the Joint Artificial Intelligence Center (JAIC) as the acting lead of their Test and Assessment group in 2019, and returned to the Naval Research Laboratory (NRL) Space Technology Division's Robotics and Machine Learning Section where she developed assurance cases and supported the development of verification and testing tools for autonomous systems. She designed the Payload Mission Manager software component for DARPA's Robotic Servicing of Geosynchronous Satellites (RSGS) program, which integrates the payload fault management system with operator generated automated, supervised, and fully autonomous behavior scripts. She served as the NRL RSGS Algorithms lead from 2014-2017 and the de facto Fault Management lead from 2015-2019. Before her arrival at NRL in 2014, she spent three years in London as the ONR Global Associate Director for Autonomy and Unmanned Systems. Prior to her term at ONR Global, Dr. Redfield worked at the Naval Surface Warfare Center, Panama City Division in Florida where she worked on heterogeneous teams of autonomous maritime vehicles and led the development of a new architecture, enabling the simulation and testing of a variety of arbitration mechanisms to control teams of vehicles.

作者簡介(中文翻譯)

西根·瑞德菲爾德是美國海軍研究實驗室（NRL）自主系統研究實驗室的主任。她曾擔任IEEE自主系統驗證技術委員會（TC-VAS）和機器人及自動化系統性能評估與基準測試技術委員會（TC-PEBRAS）的共同主席。她共同撰寫了第一個IEEE RAS標準1872-2015，並擔任標準1872.1-2024《機器人任務表示》的秘書。2019年，她被派遣至聯合人工智慧中心（JAIC），擔任其測試與評估小組的代理負責人，隨後返回海軍研究實驗室（NRL）太空技術部門的機器人與機器學習組，開發保證案例並支持自主系統的驗證與測試工具的開發。她為DARPA的靜止衛星機器人服務計畫（RSGS）設計了有效載荷任務管理軟體組件，該組件將有效載荷故障管理系統與操作員生成的自動化、監督和完全自主的行為腳本整合在一起。她在2014年至2017年間擔任NRL RSGS算法負責人，並在2015年至2019年間擔任事實上的故障管理負責人。在2014年加入NRL之前，她在倫敦擔任ONR全球自主與無人系統的副主任，任期三年。在ONR全球任職之前，瑞德菲爾德博士曾在佛羅里達州巴拿馬城的海軍水面作戰中心工作，參與自主海洋車輛的異質團隊合作，並主導開發一種新架構，使各種仲裁機制的模擬與測試得以控制車輛團隊。

類似商品