1. Introduction
In the Crisis Management (CM) domain, practitioners need to be prepared for the unexpected: based on past experience and the local incidents they had to deal with, they develop a feeling for the things that did not go so well. As with most incidents many lives are involved, they are continuously looking for solutions to improve their response and preparedness. Their aim is to be more efficient and effective, save more lives, against fewer costs. Some of the ways to achieve this are to improve existing CM procedures, like media communication. Other possibilities are to introduce new software applications or hardware to remedy current shortcomings, such as introducing a Common Operational Picture application when the situational awareness is insufficient. But how does one really know that such a solution improves things? By trying them out in a Trial, a well-controlled experiment carried out under observation.
The DRIVER+ project aims at helping CM practitioners by providing them with the 'tools' to properly evaluate new solutions during a Trail:
- The Portfolio of Solutions (PoS), an online web application where the CM practitioner can find out about existing solutions, and user experiences.
- A Lessons Learned Framework (LLF), an online web application to share experiences among practitioners.
- The Trial Guidance Methodology (TGM, deliverable D922.41 - Trial guidance methodology and guidance tool specifications (version 2)) and Trial Guidance Tool (TGT), a step-by-step approach that guides CM practitioners through all the steps needed to prepare and run a Trial.
- The Test-bed infrastructure, a suite of software tools and services that support practitioners preparing and executing a Trial or training exercise
This document is all about the latter, the Test-bed infrastructure, but since it strongly relates to the TGM, a brief summary of the TGM is presented first.
Within the DRIVER+ project, a Trial Guidance Methodology (TGM, deliverable D922.41 - Trial guidance methodology and guidance tool specifications (version 2)) and tools are developed to help resolve the needs of practitioners through a systematic and pragmatic approach. The Test-bed infrastructure is a suite of software tools and services that enables solutions and simulators in the Crisis Management (CM) domain to easily exchange information (see Figure 3). This allows end-users in the CM domain to Trial solutions, and see if they address their gaps. Additionally,it can be used to support training exercises as well.
The simplified DRIVER+ TGM process to Trial solutions is like this:
PREPARATION PHASE:
- The operational issues practitioners experience are matched with one or more of the well-known crisis management gaps, as experienced by many of their colleagues, e.g. How to get a real-time and dynamic overview of the position of all personnel?.
- These gaps, in turn, are still too generic to address, and are made more specific, leading to so-called 'research questions', e.g. Which Situational Awareness-increasing solution fits best with our mode of operation?
- Existing solutions are reviewed and selected. Some solutions can even be tried out standalone using the Test-bed and an existing mini-scenario.
- Based on the selected solutions, gaps and research questions, a data collection plan is developed: what kind of data does the Trial need to generate to enable a valid evaluation at the end of the Trial.
- A scenario is developed that puts these solutions to the test. This includes selecting and connecting simulators of a fictive incident, e.g. a flooding simulator, and other simulators to create a realistic environment, e.g. a traffic simulator.
- Selected solutions and existing legacy systems are connected to the Test-bed, so they can receive input from the simulators as well as each other.
EXECUTION PHASE:
The Test-bed is setup, all simulators and solutions are connected, data collection is in place, and a scenario can be run (executed). Before the actual Trial, several dry runs are performed, partially for testing the setup, and partially for training the participants in using the solutions. This phase ends when the Trial is executed and observed.
EVALUATION PHASE:
Based on the recorded observations and collected data (screenshots of running applications, messages sent, actions and decisions taken) during the Trial, the solutions are evaluated with the participants, leading to a good appreciation of how the selected solutions have contributed to solving the problems.
1.1 Aim of this Document
This document aims at providing the CM practitioners with a clear understanding of what the Test-bed is, and what it can do for his or her organisation. Especially chapter 1 and 2 will be of interest to them, and for the more technical oriented CM practitioner, also chapter 3. Chapter 3 and 4 are concerned with the technical setup, and are directed at developers or system administrators that need to connect new solutions or setup and maintain a Test-bed.
1.2 Scope of the document
This document limits its scope to the core Test-bed design, more specifically, the design of the Test-bed's reference implementation, which is an implementation of the Test-bed specification (see Figure 3): it thereby provides an overview of the most important components of the Test-bed, how they work together, and how they can be used by different stakeholders.
The CM solutions and simulators that supplement the Test-bed, however, are not part of the Test-bed. The simulators' function is to simulate an incident, and the reactions that may occur in a real world, since we cannot unleash incidents like a flooding and earthquakes on the real world. The solutions are the actual tools that are trialled and evaluated, and measured whether they actually do solve a CM gap. These solutions are fed with the simulator's output, and perhaps the output of other solutions, so end users can observe and evaluate their contribution during a fictive incident.
1.3 Organisation of the Document
This is a live document, and the latest version can always be found online at https://driver-eu.github.io/test-bed-design. It is organised as follows:
For reviewers, please refer to Annex 2 Change history to see what has changed since the previous version. In order to quickly understand the Test-bed, have a look at the 5 minutes animation at https://vimeo.com/299680658.
The chapter, Test-bed for Trial owners, is specifically aimed at Trial owners and practitioners, and discusses the functionality the Test-bed offers to them. And also, what it does not offer.
The Test-bed description provides a general overview of the Test-bed reference implementation and its building blocks. It is a more technical chapter aimed at Test-bed users with a (limited) technical background.
In the chapter, Test-bed for developers and sysops, the technical side of the Test-bed is explained in more detail. Specifically: how to manage a Test-bed as a sysop, or how to connect a new solution or simulator as a developer. To learn more about the design principles underlying the Test-bed, the reader is referred to Annex 3 Test-bed design.
1.4 What's new
Version 1 (2018-03-31)
This is deliverable D923.21, describing the first version of the Test-bed reference implementation. It matches the release version of the software applications as available on GitHub: github.com/DRIVER-EU. Two future releases of this document are planned, v2 and v3, and the major changes between the different documents will be described here. In the mean time, this online documentation is continuously updated, so it matches the current state of the Test-bed.
Version 2 (2018-11-30)
This is deliverable D923.22, describing the second version of the Test-bed reference implementation v2. All changes with respect to the previous version are documented in Annex 2 Change history.
Version 3 (2019-08-31)
This is deliverable D923.23, describing the third and final version of the Test-bed reference implementation v3. All changes with respect to the previous version are documented in Annex 2 Change history.