Our client was the Oil & Gas division of a global engineering firm whose CEO was looking to achieve a 10% improvement in engineering productivity within 12 months.
Nerve™ combined data from multiple systems and applied proprietary machine learning techniques to explore - and explain - the impact of people and process on the performance of project engineering teams distributed across over 119 sites in 29 countries, with 47% of projects split across at least two sites.
We looked at three years of data across the divisions of three P&Ls, accounting for nearly 70% of all engineering hours. We sourced data from over 12 disparate systems that had never been linked before, including:
- HR data; vouchering data from time management systems
- Drawing revisions from CAD
- Bill of materials (BOM) structure from PLM
- Communications data including email, calendar, and messaging
- Supplier data from ERP
- Knowledge Management tools
Due to the disparate nature of multiple latent data sets we used clustering, regression and combinatorial optimisation techniques to identify and explain a set of (non obvious) factors impacting engineering productivity.
We then used sensitivity analysis techniques to define the optimal change interventions and evaluated the trade off between different metrics (time, cost and quality) and finally used the continuing dataflow to track the impact of these mitigating initiatives over time.
From these factors, Nerve™ could prove that team design and process fragmentation were significant drivers of performance.
We identified relationships between specific locations (such as Italy and India) that drove performance and could easily be acted on. We also identified the impact of individual utilisation levels on overall team performance and demonstrated how the current working practices and targets were counter to optimal project performance.
From these factors, we found a total productivity gain of 22%, from the associated interventions the client captured 14% (worth $35m) within 12 months. These interventions were tracked through a live deployment of Nerve™ that continuously ran the analytics on the ongoing data flow and provided leadership with a continuously updated view on performance of their organisation.