In the SIITS project, we focus on developing, among other things, digital tools for interaction, visualization and analysis of risk and vulnerabilities, as well as processes for risk management. Risk visualization aids risk communication, an important part of risk management. There are several, but not universally accepted, ways of communicating risks to the stakeholders. Some traditionally used tools in this area are risk matrices, F-N curves, probability distributions, etc. Successful risk communication tools should facilitate the realisation of intended goals of the communicative actions. The US Environmental Protection Agency developed the well-known Seven Cardinal Rules of Risk Communication (see figure). Their general idea is that the stakeholders should be involved in almost all the activities of risk management through a two-way dialogue. These principles remain the best practices for risk communication till date.
In the SIITS-context, we look towards the new emerging risk visualization methods and tools that should be built on the sound principles of risk communication. Once deployed, these tools would assist in visualizing the risks around mobility by combining Information and Communication Technologies (ICTs). These tools have an important for catering to the twin challenges of describing and as well as communicating vulnerabilities & risks of the future mobility system. Some of these challenges are described below:
1. Complex and dynamic risk picture
The mobility system is exposed to different types of risks depending on geography, demography, environment, transport mode, etc. For instance, the complex Norwegian peninsula has several geographical features such as mountains, plains, costs, fjord, etc. These give rise to different types of risk exposures, challenges, and vulnerabilities. While these have posed as challenges for the traditional mobility systems as well, the changing nature of interactions is what makes the future of mobility particularly complex. Going into the future, we will see more and more dynamic interactions between vehicle-to-vehicle (V2V), vehicle-to-infrastructure (V2I), vehicle-to-pedestrians (V2P), etc., and less of the traditional human-to-human. Consequently, the future mobility system would be tightly interconnected. It may not be possible to examine a component in isolation since its future state would be partly, but strongly, determined by its interactions with other components and environment2. This produces a very complex risk picture that is changing in both space and time. Capturing the system’s state using the static and traditional risk modelling tools, that rely on separability, is insufficient and can hide underlying system uncertainties.
2. Uncertain system interactions and boundaries
The future of mobility emerges through an intersection of several systems and sub-systems. These systems use different technologies and there is a continuous inflow-outflow of data (through messages and signals) across their interfaces. Vehicle OEMs, infrastructure technology providers, smart grids, telecommunication system, service platform providers, are some of these sub-systems. In this tightly coupled technological landscape, it is challenging to define the system boundaries that are required for ascertaining the system’s description, boundary conditions and scope. The inter-system interactions at the interfaces are particularly vulnerable. For system's risk visualization, comprehending the bottlenecks, weaknesses and their likely accidental paths is quite challenging. The static risk models and tools have limited sophistication to capture such details.
3. The mobility system is prone to systemic risks
This is a new class of risk that is emerging in our world. Systemic risk refers to the risk of breakdowns in an entire system, as opposed to breakdowns in individual parts or components, and is evidenced by co-movements (correlation) among most or all parts 1. Systemic risks are particularly significant for complex systems, such as the future’s mobility. A computer virus in the traffic control station with the potential to cause malfunctioning traffic lights leading to a chain of accidents is a simplified example of systemic risk. As against the traditional risks that are somewhat bounded (e.g., fires remaining localized), the consequences of systemic risk have the possibility to spread throughout the strongly connected mobility system. The fact that even a small unnoticed initiating event in the sub-system can trigger such systemic accidents poses a real challenge to apprehend it in advance. The mechanism of consequence spreading is complex and non-linear; as against the ‘domino effect’ typically used in risk models. Capturing, modelling and communicating this risk, while being extremely important, is a definite challenge that the risk & safety community is dedicated to solving.
4. Variety of stakeholders
The future mobility could be incrementally de-centralized. This means that different participating stakeholders could impose varying level of influence on the system to achieve their objectives. For example, the regulatory authorities have vested interests to impose stricter data privacy guidelines while the autonomous technology developers would prefer greater access to user data to improve the performance of their auto-navigation system. The general users would, in turn, be conscious about their personal data being shared with the insurance companies and so on. This is a complex stakeholder landscape involved in the risk management of different aspects (economic, functional, social, environmental, etc.). Each actor expects a different set of goals, values, and utilities from this system. They would perceive the system differently, often irrespective of the common system description provided. Additionally, the stakeholders would be also interacting with each other. For example, the IT experts interacting with traffic operators, power suppliers, etc. This means that the visualization tool, while serving as a common platform for providing the system description, should also communicate stakeholder-specific information for identifying risks inherent in these interactions. Moreover, the information itself should be visualized and communicated in a clear, understandable, and unbiased manner. Achieving all these goals and finding the right balance between common- and stakeholder-specific risk information is one of the main challenges we face for the visualization tool.
5. Limitations of traditional risk communication methods
Risk communication is the step that determines the success of the risk management process. It is crucial that the risk knowledge generated is communicated to the stakeholders appropriately. Traditional risk communication has been a one-way & top-down approach, from experts to the general public. It has been conducted in centralized manner where knowledge is owned and disseminated to the world by some authoritative experts. It is based on the rational actor assumption that says that providing more accurate information will garner a more rational behavior from the receivers. However, even timely and accurate information may frequently lead to risk ‘inappropriate’ behavior. It has now been realized that risk communication needs to be a participative and dialogue driven activity. This is easier said than done, particularly in the mobility context. The ever-increasing data flows, complicated system modelling, and expert's jargons means that several stakeholders may feel left out or confused. They may even lose confidence and trust in the system. To avoid this, we would need those methods/tools that are able to intuitively communicate the analysis’ results to the decision-makers (who may lack the same level of technical expertise).
As we know, the first step in risk management of a safety-critical system is to understand and describe it. From the above discussion, clearly this is not so simple. Describing the changing mobility landscape is quite challenging as is communicating it. This is largely due to its scale, dynamicity, interconnectedness, and complexity. Even the new technologies underpinning our future mobility, while solving several important challenges from past experiences, also open opportunities to exploit safety/security gaps that didn’t exist before. For instance, the future mobility could be prone to increased cybersecurity risks. This complicates the need for risk monitoring of the system. The future mobility system will also have to be incrementally robust against malicious attacks (e.g., state-sponsors cyber-attacks, terrorist threats, etc.) and anti-fragile against local disruptions and natural calamities. The traditional risk management methodology based on analyzing the system modelled as a ‘sum of its parts’ proves insufficient. Complex mobility system demands a holistic system’s approach to risk management. We, at Proactima, plan to build and deploy digital ICT tools & techniques (backed by ML/AI) for this. These modern techniques are proving increasingly useful for managing complex systems, simulate stress scenarios and generate customized insights for different stakeholders. Such digital tools can also handle large data sets for analyzing trends, handle non-linearities, interconnectedness and determine the causes and consequences of threats (intentional/unintentional) to the system. It is also important that these tools are developed by combining the data-driven knowledge with the practical operational experiences, knowledge about risk assessment, stakeholder expectations and measures. While combining these two aspects is also a challenge, but such an approach will make the risk visualization tools truly balanced and rational. Proactima has a clear advantage here given its expertise in the risk domain as well as in building digital tools in-house. Overall, we see a true potential for these novel tools/techniques in assisting towards a safer and sustainable mobile future through the SIITS project.
1 G. Kaufman and K. E. Scott, “What Is Systemic Risk, and Do Bank Regulators Retard or Contribute to It?” The Independent Review 7, no. 3 (2003): 371–91
2 Gershenson, Carlos. "Improving Urban Mobility by Understanding its Complexity." arXiv preprint arXiv:1603.04267 (2016).