Kooperativer Bibliotheksverbund

UID:

Format: 1 online resource (173 pages)

Edition: 1st ed. 2024.

ISBN: 9783031558177

Series Statement: Studies in Computational Intelligence, 1150

Content: The fast development of intelligent technologies in the previous years, their boost during the COVID crisis, and their huge acceleration with the proliferation of the Large Language Models has led industry, governance, legislators and regulators to race to achieve production control, striving to define regional and global regulatory frameworks, capable of guaranteeing their safe deployment. In an evolving and uneven regulatory context, balancing fast technological development and the safe and ethical use of these new disruptive technologies, with an expected total economic impact of 4.4 trillion dollars annually, becomes presently the biggest challenge. The present book addresses the fundamental role played by the three milestones- Benchmarking, Standardization, and Certification that are part of the developmental loop of all products in developed economies. The essential role played by these milestones and the importance of defining accurate general and specific metrics is addressed in this book with particular case studies in the domains of Robotics and AI-.

Note: On the relationship between Benchmarking, Standardisation and Certification in Robotics and AI -- Confiance.ai Program, Software Engineering for Trustworthy AI -- Trust in Robot Benchmarking and Benchmarking for Trustworthy Robots -- Trustworthy Artificial Intelligence in the Context of Metrology -- Performance Modelling of Autonomous Mobile Robot Software Components -- Towards Standardization of Interaction Design: Cognitive Developmental Robotics Revisited -- KPI-driven Energy Monitoring Dashboards for Robotic Performance -- Co-Regulation and AI Innovation -- Prolegomenon for a Family of Theorems Regarding Trustworthiness in Autonomous Artificial Agents -- The Quest for an AI Ethics: Between Benevolence and Greed.

Additional Edition: Print version: Aldinhas Ferreira, Maria Isabel Producing Artificial Intelligent Systems Cham : Springer International Publishing AG,c2024 ISBN 9783031558160

Language: English

DOI: 10.1007/978-3-031-55817-7

URL: Volltext  (URL des Erstveröffentlichers)

UID:

Format: 1 online resource (173 pages)

Edition: 1st ed.

ISBN: 9783031558177

Series Statement: Studies in Computational Intelligence Series ; v.1150

Note: Intro -- Preface -- Contents -- On the Relationship Between Benchmarking, Standards and Certification in Robotics and AI -- 1 Introduction -- 2 Benchmarks -- 3 Standards -- 4 Linking Benchmarks, Standards and Certification -- 4.1 Regulation and Related Standards in Robotics -- 4.2 Regulation and Related Standards in AI -- 5 Discussion and Conclusions -- References -- Confiance.ai Program Software Engineering for a Trustworthy AI -- 1 Introduction -- 2 Reminders About AI -- 2.1 What is AI? -- 2.2 Technologies -- 2.3 Applications -- 3 Weaknesses of AI -- 3.1 Unknown Unknown, Data Obsolescence -- 3.2 Data Bias -- 3.3 Lack of Robustness, Adversarial Attacks -- 4 Regulation -- 5 The Confiance.ai Program -- 5.1 Aim -- 5.2 Consortium -- 5.3 Structure of the Program -- 5.4 Use Cases -- 5.5 Deliverables -- 5.6 An Open Consortium -- 6 Results -- 6.1 Robustness Assessment -- 6.2 Robustness by Design -- 6.3 Explainability -- 7 Conclusion -- References -- Trust in Robot Benchmarking and Benchmarking for Trustworthy Robots -- 1 Introduction -- 2 Progressive and Iterative Robot Benchmarking -- 2.1 Dataset-Based Evaluation -- 2.2 Remote Evaluation -- 2.3 Functional Benchmark -- 2.4 Task Benchmark -- 2.5 Completing the Cycle -- 3 Developing Benchmarks for Trustworthy Robots -- 3.1 Datasets -- 3.2 Benchmark Definition for Field Campaigns -- 3.3 Social Acceptance -- 4 Conclusions -- References -- Trustworthy Artificial Intelligence in the Context of Metrology -- 1 Introduction -- 2 Research on TAI at NPL -- 2.1 Explainable AI -- 2.2 A Metrology Framework for Uncertainty Evaluation in Machine Learning -- 2.3 Good Practice in Training Set Preparation for Marine Navigation Systems -- 3 Certification of AI Systems -- 4 Concluding Remarks -- References -- Performance Modelling of Autonomous Mobile Robot Software Components -- 1 Introduction. , 2 Background and Literature Review -- 2.1 Benchmarking Complete Robot Systems -- 2.2 Benchmarking 2D SLAM and Localization Methods -- 2.3 Benchmarking Autonomous Navigation Methods -- 3 Performance Modelling Methodology -- 3.1 Formal Definition -- 3.2 Performance Model Composition -- 4 Component Performance Modelling -- 4.1 Experimental Setup and Protocol -- 4.2 Environment Features -- 4.3 Performance Metrics -- 5 Use Case: SLAM Performance Models -- 6 Use Case: Local Planning Performance Models -- 7 Use Case: Autonomous Navigation Performance Models -- 8 Conclusions and Future Work -- References -- Towards Standardization of Interaction Design: Cognitive Developmental Robotics Revisited -- 1 Cognitive Developmental Robotics and Its Philosophical Background -- 2 Revisiting the Physical Embodiment and Social Interaction -- 3 The Meaning of Physical Body (Ecological Self) -- 4 Between Embodiment and Social Interaction (Interpersonal Self) -- 5 Social Interactions (Social Self) -- 6 Discussion -- References -- KPI-Driven Metric Acquisition Methodology with a Energy-Centric Robotic Performance Case Study -- 1 Introduction -- 2 KPI-Driven Methodologies and Applications -- 3 Energy Condition Monitoring for Robotic Performance -- 3.1 Approach Overview -- 3.2 Proposed Motivating Factors -- 3.3 Experimental Test Bed -- 3.4 Data Acquisition -- 3.5 Network Gateway -- 3.6 Data Management -- 3.7 Data Visualisation -- 3.8 Data-Flow Between Layers -- 4 Energy-Targeted Visualisation -- 4.1 KPIs and Metrics -- 4.2 Dashboard Designs -- 5 Discussion -- 6 Conclusion and Future Works -- References -- Co-regulation and AI-Innovation: Principles for a Sustainable Framework Fostering Innovation and Acceptance of AI -- 1 Introduction -- 2 Co-regulation Under the Proposed AI-Act -- 3 Formal Aspects: Regulatory Quality -- 3.1 EU Perspective. , 3.2 AI-Act and Co-regulatory Quality -- 4 Material Aspects: Innovators Perspective -- 4.1 Legal Certainty, Stringency, and Flexibility in Theory -- 4.2 Legal Certainty, Stringency, and Flexibility in the AI-Act -- 5 Material Aspects: Consumers Perspective -- 5.1 Risk and Trust in Theory -- 5.2 Risk and Trust in the AI-Act -- 6 Conclusions and Recommendations -- References -- Prolegomenon for a Family of Theorems Regarding Trustworthiness in Autonomous Artificial Agents -- 1 Introduction -- 2 What Autonomy in Artificial Agents Is Not -- 3 The Cycle of Autonomy and Its Key Factors -- 4 The Trust Relation script upper TmathcalT -- 5 A Worrisome Case Study -- 6 The Future -- References -- The Quest for an AI Ethics: Between Benevolence and Greed -- 1 An Ontological Paradigm Shift: When Tools Become Autonomous Entities -- 2 The Existence-Centered Ethical Framework: The Imperative of Responsibility -- 3 The Metrics for Responsible AI -- 4 Regulatory Frameworks and Governance -- 5 The Myth of Cooperation -- References.

Additional Edition: Print version: Aldinhas Ferreira, Maria Isabel Producing Artificial Intelligent Systems Cham : Springer International Publishing AG,c2024 ISBN 9783031558160

Language: English

UID:

Format: xii, 167 Seiten : , Illustrationen, Diagramme (vorwiegend farbig).

ISBN: 978-3-031-55816-0

Series Statement: Studies in computational intelligence volume 1150

Additional Edition: Erscheint auch als Online-Ausgabe ISBN 978-3-031-55817-7

Language: English

Subjects: Computer Science

UID:

Format: XII, 167 p. 46 illus., 38 illus. in color. , online resource.

Edition: 1st ed. 2024.

ISBN: 9783031558177

Series Statement: Studies in Computational Intelligence, 1150

Content: The fast development of intelligent technologies in the previous years, their boost during the COVID crisis, and their huge acceleration with the proliferation of the Large Language Models has led industry, governance, legislators and regulators to race to achieve production control, striving to define regional and global regulatory frameworks, capable of guaranteeing their safe deployment. In an evolving and uneven regulatory context, balancing fast technological development and the safe and ethical use of these new disruptive technologies, with an expected total economic impact of 4.4 trillion dollars annually, becomes presently the biggest challenge. The present book addresses the fundamental role played by the three milestones- Benchmarking, Standardization, and Certification that are part of the developmental loop of all products in developed economies. The essential role played by these milestones and the importance of defining accurate general and specific metrics is addressed in this book with particular case studies in the domains of Robotics and AI-.

Note: On the relationship between Benchmarking, Standardisation and Certification in Robotics and AI -- Confiance.ai Program, Software Engineering for Trustworthy AI -- Trust in Robot Benchmarking and Benchmarking for Trustworthy Robots -- Trustworthy Artificial Intelligence in the Context of Metrology -- Performance Modelling of Autonomous Mobile Robot Software Components -- Towards Standardization of Interaction Design: Cognitive Developmental Robotics Revisited -- KPI-driven Energy Monitoring Dashboards for Robotic Performance -- Co-Regulation and AI Innovation -- Prolegomenon for a Family of Theorems Regarding Trustworthiness in Autonomous Artificial Agents -- The Quest for an AI Ethics: Between Benevolence and Greed.

In: Springer Nature eBook

Additional Edition: Printed edition: ISBN 9783031558160

Additional Edition: Printed edition: ISBN 9783031558184

Additional Edition: Printed edition: ISBN 9783031558191

Language: English

DOI: 10.1007/978-3-031-55817-7

URL: https://doi.org/10.1007/978-3-031-55817-7