Office for Product Safety Standards supports Queen Mary research on novel methods of risk assessment
24 November 2020
The Government's Office for Product Safety Standards (OPSS) is funding a research project at Queen Mary to examine novel methods of product risk assessment.
It is vital that consumers are protected against the possible dangers presented by unsafe products that they might buy and use.
Product risk assessment is undertaken by relevant Market Surveillance Authorities to determine the level of risk presented by a product and, accordingly, whether action is required to protect consumers from that risk.
The new project will build on the work of Queen Mary's Risk and Information Management (RIM) Research Group, led by Professor Norman Fenton, which uses a technique called Bayesian networks to produce probabilistic risk assessments by combining causal knowledge and data. The aim is to develop a new Bayesian network model for a variety of household products. The model will incorporate behavioural evidence about risk attribution.
A prototype model has already been produced by Joshua Hunte, a PhD student supervised by Norman and Professor Martin Neil in the RIM Group. It can also be used to complement existing methods of risk assessment that are currently employed in OPSS.
The new grant will run from October 2020 to March 2021 and primarily provides additional support for Hunte to further develop the model. It will also enable the research team to undertake online experiments to empirically investigate consumer's perceptions of product risk.
Graham Russell, Chief Executive of OPSS said: "We are excited to formalize our relationship with Queen Mary's Risk and Information Management Research Group. With its emphasis on full quantification of uncertainty, this project has the potential to make significant impact with respect to supporting understanding of, and improving, the quantification of risk in product safety risk assessment."
More information about Queen Mary's Risk and Information Management Research Group can be found here.
Updated by: Michal Filus