Our project addresses Challenge 2: Data & Digitalisation, along with Challenge 1: Whole System Integration.
SGN is moving towards data-led operational management to drive improved safety and productivity. Our data strategy is based on the concept of building digital twins of operations as they actually occur in the field. We have already deployed FYLD to 68,000 jobs over the past 18 months, which has vastly improved our data set. Worker buy-in is critical to achieving high-quality data sets. Therefore, our goal is to deliver more wins that benefit our workers, SGN as an organisation and our customers.
SGN has shown its capability to adapt to new, data-led ways of working in its initial deployment of FYLD. The entire business has deployed FYLD to proactively manage fatigue whilst concurrently driving down network operation costs. In addition, 750 operatives in its repair, replacements and connections business unit now use FYLD as their primary work management platform. The next logical step in our data transformation and AI journey is to harness FYLD data to predict job sites with a high risk of safety incidents or injuries and drive more productive field force operations. 
FYLD has developed significant expertise in terms of AI and building deployable machine learning models. Additionally, we have a world-renowned AI expert Distinguished Professor & Executive Director of Data Science at the University of Technology, Sydney Dr Fang Chen as a core advisor to the business. Dr Chen has a proven track record of deploying predictive analytics models to the utilities industry and has collaborated extensively on the commercialisation of such systems with the CEO of FYLD, Shelley Copsey, during their years working together at Australia's national research agency CSIRO. 
SGN and FYLD innovation partnership is two years old, with a further three-year term being agreed. We are well placed to deliver this project together, given our past success in safety and productivity outcomes for SGN's field force operations.
FYLD has gained substantial traction in the utilities industry supply chain, including with Morrisons, Ferrovial, Lanes Group and Galliford Try, as well as being part of the HS2 supply chain after successfully securing a place on its coveted Accelerator Program. Potential users of our innovation include the SGN supply chain and beyond, given the safety and productivity outcomes FYLD delivers today, combined with the step-change in worker safety outcomes we are targeting with this project.

Problem Bring Solved

Utility companies face the problem of plateauing safety performance. We aim to make a step-change in worksite safety through leveraging operational site data and Artificial Intelligence (AI) while simultaneously achieving productivity gains. The project forms part of the building blocks of SGN's data strategy and deployment of next-generation, user-driven, digital products. 
The safety of worksites across the utilities industry has plateaued for five years, and workplace deaths actually increased in 2021. Sixty members of the public also lost their lives due to workplace incidents. These trends are concerning, particularly noting the impact of Covid on workforce management and a shift to remote operations. In the 12 months to March 2021, SGN recorded 2,506 field workforce incidents and near misses. Fieldworkers operate in dangerous environments daily, with the potential for life-changing consequences when hazards are not thoroughly identified and controlled. Shutting job sites down impacts productivity and causes job backlogs to rise. In addition, network operators face the damaging consequences of poor customer satisfaction.
FYLD has already made significant steps to improve the safety and productivity of field teams by capturing substantial volumes of unstructured data (voice, video, imagery, text) about workforce operations and using this data to generate AI driven point in time risk assessments. The next phase of development is to harness that data for predictive analytics about safety and operational events, enabling SGN to make proactive interventions. It will involve the redesign of major operational processes with data led cost-benefit analyses.
This project is the first step towards that vision. During the Discovery Phase, we will:

• Identify all relevant data sets held that could contribute towards predictive capability for safety interventions, including proprietary data (job hazards, controls, incidents, operational job delivery), other network data (to be identified), and third-party data (weather, geospatial);
• Apply Machine Learning techniques to merge disparate data sets where possible;
• Seek to identify correlational or causal factors in respect of safety incidents and other data held; and
• Build an MVP model based on our learnings to validate whether we can deploy a predictive safety intervention into production systems.

Our goal is to build an operating system that lets SGN manage safety through a proactive intervention system that could genuinely lead to a zero-harm outcome for workers. Additionally, SGN will continue to build confidence in its ability to deploy predictive models to drive large scale operational efficiencies.