PhD Throwback Series

The motive

This year is very special for me as it marks 5 years since I moved to the UK to complete my PhD studies, start my professional career and begin a new life in the country I love the most, the UK. I have been thinking about it for a while now, so the thoughts finally turned into words; I decided to celebrate this important anniversary by publishing a series of articles consolidating my research as a tribute to my parents, my friends, my colleagues and myself.

STEM has been my life and my passion. I am in the unique position of being able to draw on both academic and industrial experience, take the best from each and contribute holistically. I have been leading high-profile research and industry projects for more than a decade. In 2018, I made the life-changing decision to join industry and expand my horizons after having achieved maximum research performance and a high standard of academic reputation. I have gone strength to strength and worked my way up to driving the R&D Function for a major company in the UK and globally.

The series

In the first article of this series, I will describe the big picture of my research and later down the line I will dig deeper into the topics I have been examining over the years. But first things first: during my PhD, I developed a multivariable indicator for the quantitative assessment of system performance called RiskSOAP. My idea originated from Computer Science and Pattern Matching; a mechanism for checking a given sequence of tokens for the presence of the constituents of some pattern as shown in Figure 1. In a similar manner, my approach was based on the comparison of work-as-imagined/designed and work-as-done.

Figure 1. The process of pattern matching

Parts of my thesis were published in 7 journals, several conferences internationally and presented at the Summer School on Information Systems for Crisis Response and Management at Campus Vesta (Belgium), a training centre for emergency services dealing with public safety and security.

Part #1 - The big picture: socio-technical systems, safety and risk awareness

Engineering systems are a typical example of complex socio-technical systems, as they comprise of technical, social and governance aspects. While their mission is to be in the service of people by offering them high quality services and infrastructure, they are also identified as safety-critical because a failure or malfunction may result in death or serious injury to people. Practice has shown that even in meticulously engineered systems incidents or accidents are inevitable. Thus, it is apparent that there is a need for tools to support safety-driven design and accident prevention mechanisms for complex engineering systems.

In complex socio-technical system, safety is treated as an emergent property and a control problem for the reason that to be maintained within acceptable levels, adequate feedback processes should be designed and ‘installed’ within the system. Safety depends on the enforcement of control actions (see Figure 2 blue triangle) on the behaviour of system components (e.g. agent; actuator etc.), the parts (i.e. the loop) and their interactions (i.e. arrows) as depicted in Figure 2.

Figure 2 represents a control loop, which is the elementary part of the complex socio-technical system model. Socio-technical systems comprise of many control loops, the number and the interdependencies of which depend on the complexity and the composition of the system.

Figure 2. The components of a typical control loop and the (inter)actions between them

A key factor in system safety is risk Situational Awareness (SA). Systems have a capability to perceive and comprehend threats and vulnerabilities and project what they may entail for system safety. The risk SA – as I call it in my PhD thesis – is the individual SA of a system agent (i.e. decision maker/ process controller) and refers to being conscious of the threats and vulnerabilities that may lead to system accidents if they go unnoticed and untreated. There is a feedback action (see Figure 2 yellow triangle) that contributes to the risk SA of each agent so that he/she/it (‘it’ for automation) enforces the right control action to tackle threats and vulnerabilities.

Aside from the risk SA, in my PhD thesis I introduce the term risk distributed SA. The distinction between SA and DSA lies in the fact that the former can be found in individuals (i.e. held by the agent of a single loop), while the latter is an emergent property of the socio-technical system (i.e. held by the system, not by individuals). Risk DSA indicates that each agent on the one hand may have a detailed picture of the threats and vulnerabilities of the part he controls, but on the other hand retains a partial view of the issues threatening the entire system.

Based on the above, the problem that this thesis investigates is the degree to which various elements of a complex socio-technical system affect the system’s risk DSA. The goal is to facilitate engineers and designers in choosing the system design that is likely to maximise or at least enhance risk DSA. An important contribution of this thesis is the proposition of the RiskSOAP indicator that provides a quantitative reflection of a system's safety performance, it proves the positive correlation between safety and risk SA and supports the assessment of the risk DSA.

RiskSOAP is a comparison-based methodology that goes through three stages; that is, (1) determine the desired composition of the system, then (2) identify the as-is configuration, and finally (3) employ a comparative strategy in order to depict the distance between the compared units. To obtain that, the RiskSOAP methodology is founded on three existing approaches: (a) the System-Theoretic Process Analysis (STPA), (b) the Early Warning Sign Analysis based on STPA (EWaSAP), and (c) the Rogers-Tanimoto dissimilarity measure for binary data.

To demonstrate how the RiskSOAP methodology works in practice, RiskSOAP is applied to three real socio-technical systems: a robotic installation, an aviation system, and a road tunnel. Also, the RiskSOAP indicator is assessed against quality standards and criteria. RiskSOAP is a global indicator; that is, if for instance the main mission of the system is not safety, but profit, then the behaviour of the system can be modelled and simulated by economic or econometric models instead of accident models. Overall, this new indicator can drive the design and development of reactive and resilient complex socio-technical systems.

The big picture

RiskSOAP and the underlying systems theory inspired me later on in my postdoctoral research in the fields of healthcare (University of Cambridge, Engineering Design Centre), rail and unmanned vehicles (Imperial College London, Centre for Transport Studies), infrastructure and nuclear (Imperial College London, Centre for Systems Engineering & Innovation) and remain a valuable source of insight and innovative solutions feeding into the high-profile and challenging projects I am leading in my industrial career.

References

[1] Chatzimichailidou, M.M., Protopapas, A., & Dokas, I.M. (2014). Seven issues on Distributed Situation Awareness measurement, In Complex Socio-technical Systems.

[2] Chatzimichailidou, M.M., Katsavounis, S., & Dokas, I.M. (2015) Measuring the Situation Awareness Provision Capability in Complex Socio-technical Systems with STAMP. In STAMP Workshop.

[3] Chatzimichailidou, M.M., Stanton, N., & Dokas I.M. (2015). The concept of risk Situation Awareness provision: towards a new approach for assessing the DSA about the threats and vulnerabilities of complex socio-technical systems. Safety Science, 79, 126–138.

[4] Chatzimichailidou, M.M., & Dokas I.M. (2015). Introducing RiskSOAP to communicate the Distributed Situation Awareness of a system about safety issues: an application to a robotic system. Ergonomics, 1–37. 

[5] Chatzimichailidou, MM, & Dokas I.M. (2015). Assessing Distributed Situation Awareness in Socio-Technical Systems with RiskSOAP. 2nd International Conference on Information Systems for Crisis Response and Management in Mediterranean Countries (ISCRAM-Med).

[6] Chatzimichailidou, M.M., & Dokas I.M. (2015). The Risk Situation Awareness Provision Capability and its Degradation in the überlingen Accident Over Time. In European STAMP Workshop.

[7] Leveson, N. (2011). Engineering a safer world: systems thinking applied to safety, MIT Press.

[8] Endsley, M.R. (1995). Toward a theory of situation awareness in dynamic systems. Human Factors: The Journal of the Human Factors and Ergonomics Society, 37(1), 32–64.

[9] Stanton, N.A., Salmon, P.M., Walker, G.H., Baber, C., & Jenkins, D.P. (2005). Human factors methods: a practical guide for engineering and design, Ashgate.

[10] Dokas, I.M., Feehan, J., & Imran, S. (2013). EWaSAP: an early warning sign identification approach based on a systemic hazard analysis. Safety Science, 58, 11–26.