Data Centric Safety

by Alastair Faulkner and Mark Nicholson

Binding: (Paperback) c. 520 pages
Publisher: Elsevier; 01 edition (Mar 01, 2020)
Language: English
ISBN: 978-0-12-820790-1

Technology evolves, shaped by its use, often in unexpected ways. Products, once constrained by 'air gaps', are enabled by communications-based infrastructural technologies and their data ecosystems. Yet data is too broad a term, as it does not address its many roles within systems. It is true that historically in most systems, data is merely consumed, processed and some action performed based on predetermined criteria. In this case, data is passive, inert, requiring consumption to participate in or direct actions or activities. However, data often exhibits many degrees of freedom that include the description of functionality, performance, capability, capacity and constraint. Data may also include temporal (sequence or order) or time-based (time, rate or calendar) properties. Data has a mercurial property; it is challenging to manage and control, it has a habit of being consumed by systems that it was not produced for; by omission or by design, perhaps not to the awareness of the system designer, it is common for it to pass (often unchecked or even unwittingly) across system and organisational boundaries.

Data (in all its forms) is often unchallengeable, unverifiable, ubiquitous, unrecorded and invisible. Yet this data increasingly determines the behaviour of systems and through this behaviour our access to products (goods and services). Data may be internal or fed to systems with safety responsibility. As a result data error or omission may go undetected with potentially hazardous or catastrophic consequences. There may also be consequent damage to assets. Failure of such systems may also contribute to harm indirectly through incorrect decisions made by actors (human or computer) who rely on, or trust, these systems and the data they supply. How should we reason about Data-Centric Systems (DCS) so that our reliance on, or trust in their correct operation can be justified?

In using the term DCS, we acknowledge the ever-increasing volumes of data, in all its dimensions from size to scale and complexity. Data may be structured or unstructured. However, not all data has value to us; not all data is fit to be used in a system with safety implications, or as part of the assurance of such systems. So how should we determine what data can be used, and what it can be used for? How do we assure ourselves that data used is appropriate and has the right characteristics? How do we engineer our systems to ensure they are robust and resilient to data errors and failures?

As these DCSs grow, they experience a \textit{change-of-scale}, consuming (and potentially producing) vast quantities of data. As a result (automated) methods are required to ensure and assure the contribution of data to system safety in such systems. How further do we assure that actors using the data generated by such systems do so in the intended way and with the appropriate level of criticality?

Currently, no mature methods exist to address these issues. Careful, development of data-intensive systems will improve an organisations ability to ensure and assure the safety of such systems. Currently, guidance in this area is very immature. We address these issues in this book.

System Safety Engineering is applicable across the entire life-cycle of a product from concept to disposal. In this book, we address data safety issues relating to both physical goods and service elements of a product within a SSE framework. However, the emerging field of data-safety means that in this first edition there are aspects of data safety that we do not address.

Contents
Part I Data Centric Safety
1 Introduction (and Structure of the Book)
2 System Safety Management
3 Challenges to Systems Engineering
Part II Data Centric Fundamentals
4 Data Fundamentals
5 Data Centric Systems
6 System Context
7 System Definition
Part III Data Centric Design
8 Data Centric Architecture
9 Development
10 Acceptance and Approval
Part IV Operational Management and Maintenance
11 Operational Matters
12 Live Management and Control
Part V Incident Investigation
13 Major Incident Response
14 Investigation Management
15 DCI Investigation Methodologies
16 Incident Investigation
17 Investigation Methodology Maturity
18 Analysis as part of a DCI
19 Incident Report
Part VI Data Safety Model
20 Data Safety Model
21 Using the DSM
22 Validation
Part VII Application Areas
23 Autonomous Flight
24 Enterprise
25 Healthcare
Part VIII References