Research and development services and related consultancy services

Description

HSE are currently 18 months in to a 5-year programme of work, the Discovering Safety programme (DSP), supported by a research grant from Lloyd’s Register Foundation. The primary aim of the research programme is to understand how to access and use routine safety and health data in new ways to provide a new approach to improving performance by applying novel data science, artificial intelligence and machine learning techniques in health and safety contexts. The vision for the Discovering Safety programme is to deliver step change improvements in global health and safety performance through better analytic exploitation of routine health and safety data. The focus of this invitation to tender is for two of the construction industry use cases, the construction health and safety risk library and leading indicators for health and safety performance assessment. Phase 1 work ended with a limited piloting of the prototype tool developed in order to demonstrate proof of concept to the Discovering Safety programme board. Broad ambitions for Phase 2 With respect to Phase 2 work, the DSP team wish to: 1) Further develop the prototype tool from Phase 1, specifically, to widen out its scope to include other categories of projects, work activities and categories of risks; 2) Further pilot the tool across select parts of industry (both in the UK and internationally); and 3) Explore the viability of creating a similar tool able to serve up health and safety knowledge to a project for use by contractors (i.e. once construction works have actually started). Broad ambitions for Phase 2 In Phase 2 of the DSP, the DSP team wish to work with construction organisations more directly, to consider how their routine health and safety data could be put to better use, using the framework developed in Phase 1 as a useful reference. The aim is to create digital solutions that organisations could make use of in this context. Monitoring how a construction site is performing from a health and safety perspective typically forms an important part of its health and safety management arrangements. This tends to be done by collating agreed key performance indicator metrics derived from the various sources of routine health and safety data generated and reporting back to managers and supervisors using dashboards and scorecards. As a rule, the primary focus of such reporting tends to be around relatively simple, high level health and safety related concepts that can be easily quantified. In the main, this involves working with pre-structured data, i.e. counts, rates or categorical data. Aspects of operations captured in free text narratives, such as accident reports, safety observations, and inspection, audit and investigation findings, tend to be much less widely used in the process. The health and safety digital technology market has numerous products aimed at supporting organisations in the collection and descriptive analysis of operational KPI’s, including those linked to safety related processes and outcomes. Very often these are linked to, or form an integral part of, an organisation’s enterprise content management system. However, for most of these products, the analytics performed is largely descriptive, to support operational reporting tasks, often through the use of score cards and dashboards. The use of KPI data collected for predictive analytic purposes, to directly inform inspection regimes and other interventions, is much less common. The broad ambitions for Phase 2 work is to explore the viability of bringing together the high level thinking from Phase 1 work, along with the capabilities being developed as part of the text mining work, to look to add to the functionality of existing digital solutions. To achieve this, a clear understanding of how the data governance process works on existing projects is regarded to be key, including the sorts of routine health and safety data collected and how it is stored and made use of currently.