Jon Wackrow
London Underground
Jon is the Head of Human Factors at London Underground, a position he has held for 10 years. He is based in the organisation's Capital Programmes Delivery directorate. Here the future railway is designed, engineered and delivered through the portfolio of line upgrades and major civils programmes. Jon is responsible for interpreting the changing demands for HF and determining the most efficient way of providing solutions to the challenges they represents. This includes HF engineer accreditation activities and ongoing monitoring and development of company HF standards to ensure engineers and standard guidance together continue to deliver best value. His presentation today covers one such recent standard development. As well as a corporate element, Jon's role has always had a delivery element too and he is currently the New Tube for London HF Delivery Manager where challenges of end user integration within increasingly automated systems presents specific and new challenges for HF to solve.
Good alarm system design is based on clarity and consistency of terms and definitions, and their relationships, used to describe events that are worthy of alarming. Achieving this provides a basis, and rationale, for alarm... [ view full abstract ]
Good alarm system design is based on clarity and consistency of terms and definitions, and their relationships, used to describe events that are worthy of alarming. Achieving this provides a basis, and rationale, for alarm lists and their design; each required to facilitate good and reliable alarm handling performance.
Alarm system design guidance is either context specific, or generic. Terms and definitions, and their relationships, used in the rail industry will be specific to its requirements for alarm worthy events, and their presentation design, and they will be different to those for the water industry, for example.
This paper proposes that there is a lack of clear and consistent guidance, both within and between design documentation, which inhibits rail companies from efficiently delivering alarm system design supporting good and reliable alarm handling performance. The use of lexicon and glossary contribute to the lack of clarity and consistency of terms and definitions, and their relationships.
This paper contends that document modelling utilising visual representations of terms and their relationships, are a more effective way to determine, and present clarity and consistency around the subject of rail alarm systems.
This paper describes the work that was carried out to bring clarity and consistency to alarm terms, and their relationships, for rail, but specifically London Underground (LU) application. It includes the assessment of current key alarm system design guidance, case studies of LU alarm system developments, and findings from baselining current LU systems’ performance. By visually modelling the outputs of these activities, and thereby presenting key terms and their relationships visually, it was possible to clearly and consistently define alarm worthy events, and their key design features, ie. create a visual ontology for a rail alarm systems.
The findings are that the alarm system ontology does start to clear up the wide and confused lexicon/glossary problem for rail, specifically LU. It does this by bringing clarity to alarm terms and definitions, and their relationships.
System design is a response to a system problem and only when the problem is known and articulated can system design commence. Clearly defining a system problem leads to an improved system development process which leads to an improved system design. The alarm system ontology’s visual relationships reflect the LU context of use, and are superior to unstructured textual lexicon and glossaries by allowing the reader to understand LU alarm terms and definitions, and their relationships. Having defined the problem, the ontology can then be used to drive improved LU alarm system design to support good and reliable alarm handling performance.
The paper suggests further work that would improve the detail of the alarm system ontology, and generate empirical data about its efficacy. On the basis of the mutual dependency between ontology and alarm system development process, it is recommended that most benefit from development of the ontology will be achieved when it is developed alongside existing rail alarm system development processes.
