Simon Thorne

The Misuse of Spreadsheets in the Nuclear Fuel Industry: The Falsification of Safety Critical Data Using Spreadsheets at British Nuclear Fuels Limited BNFL

This paper considers the management, technological and human factor issues that led to the BNFL fuel rod spreadsheet data falsification incident in 1999. BNFL discovered in 1999 that some data supporting quality assurance and safety processes had been falsified by BNFL workers using spreadsheets. The implication of this finding was that some of the Mixed Oxide Fuel Pellets shipped to customers in Japan for use in nuclear reactors were of an unknown mass and quality. This meant that the use of the MOX pellets fuel produced by BNFL would introduce uncontrolled factors into the safe operation of nuclear reactors. This paper will examine the production of MOX pellets at the Sellafield site, the falsification of data and the report commissioned by HM Nuclear Inspectorate. The paper will then identify a number of managerial and technological failings that led BNFL to use spreadsheets for recording such data. Finally the paper analyses other cases of spreadsheet fraud and explores some strategies for reducing the likelihood and impact of spreadsheet errors and fraud.

M-Health Solutions to Support the National Health Service in the Diagnosis and Monitoring of Autism Spectrum Disorders in Young Children

With estimates of prevalence between 1 in 68 and 1 in 88 children [11], accurate and early identification of autism spectrum disorders (ASDs) in young children remains a pressing public health issue. In the absence of a single biomarker for ASD, however, a diagnosis is currently reached on the basis of a portfolio of evidence assembled by various health care professionals, parents and educational specialists. Studies have shown that early diagnosis and subsequent intervention are key to a favourable prognosis for children with autism. Many families, however, experience long periods of time between appointments with health care professionals, thus delaying the diagnosis and subsequent access to support and interventions. In this paper, we consider the potential role of m-health software solutions in supporting the diagnosis and ongoing monitoring of ASDs in young children. We consider their application particularly within the context of the UK’s National Health Service. This paper also presents a review of some of the current literature on user-behaviour analysis software on mobile computing devices such as tablet computers and smartphones, along with some of the emerging m-health solutions for supporting the diagnosis of ASD in children.

The Misuse of Spreadsheets in the Nuclear Fuel Industry: The Falsification of Safety Critical Nuclear Fuel Data Using Spreadsheets at British Nuclear Fuels Limited (BNFL)

This paper considers the management, technological and human factor issues that led to the BNFL fuel rod spreadsheet data falsification incident in 1999. BNFL discovered in 1999 that some data supporting quality assurance and safety processes had been falsified by BNFL workers using spreadsheets. The implication of this finding was that some of the Mixed Oxide Fuel Pellets shipped to customers in Japan for use in nuclear reactors were of an unknown mass and quality. This meant that the use of the MOX pellets fuel produced by BNFL would introduce uncontrolled factors into the safe operation of nuclear reactors. This could result in uncontrolled nuclear reactions and may have serious implications. The BNFL workers had cloned spreadsheets containing micrometer measurements and adjusted lot and batch numbers so that each spreadsheet appeared to be a genuine reflection of the quality and safety procedures put in place to ensure that the pellet dimensions, density and surface features are known and within acceptable tolerances. This paper will examine the production of MOX pellets at the Sellafield site, the falsification of data and the report commissioned by HM Nuclear Inspectorate. The paper will then identify a number of managerial and technological failings that led BNFL to use spreadsheets for recording such data.

Software Requirements Engineering in Digital Healthcare: A Case Study of the Diagnosis and Monitoring of Autism Spectrum Disorders in Children in the UK’s National Health Service

A major issue in designing digital healthcare software solutions is ensuring they meet the clinical needs and requirements of key services, as well as the expectations of various healthcare professionals. Modern software requirements engineering must be adapted to cater for this demand; we argue that traditional (and popular) requirements engineering processes – particularly in relation to the elicitation and validation of key requirements – may not be the most appropriate within the context of a multi-disciplinary team of healthcare professionals. Successful software requirements engineering is vital in ensuring that digital healthcare solutions fulfill expectations and meet the clinical needs; we thus propose that new methods of gathering requirements in the ‘third space’ are needed. This paper draws on a case study of the multi-disciplinary team of healthcare professionals involved in the diagnosis and support of autism spectrum disorders (ASD) in young children within the UK’s National Health Service (NHS). It is worth noting that, in the context of our case study, requirements engineering is an iterative process and requires the input of numerous stakeholders from often stretched and fragmented services.

Reducing Error in Spreadsheets: Example Driven Modeling Versus Traditional Programming

This article presents experimental data supporting an alternative approach to developing decision support spreadsheets using a Programming by Demonstration paradigm. This technique is coined “Example Driven Modeling” and uses example data (attribute classifications) in combination with inductive machine learning to create decision support models as an alternative to spreadsheet programming. This experiment examines whether participants can define attribute classifications (“example-giving”) satisfactorily and describe benefits and limitations this method offers through statistical analysis of the experimental results. The article then considers the wider implications of this research in traditional programming.

The potential of Example Driven Modelling for Decision Support Spreadsheets

In this paper we present a novel alternative means of modelling Decision support spreadsheets called Example Driven Modelling. The concept and rationale of the approach is discussed and experimental evidence is presented illustrating the benefits of the methodology. We demonstrate the application of EDM on a real world spreadsheet and compare the performance of the EDM model with the traditional spreadsheet. Finally we discuss the relative advantages and disadvantages of this approach and consider other areas of application in spreadsheet modelling