Machine learning for manufacturing efficiency
Case Study
Background
The Challenge
Responding to the needs of its clients, small and micro manufacturers, the software provider set out to develop a bench-marking and analytical insights service.
Their vision would allow these organisations to access and benefit from insights into their performance against the manufacturing industry, while highlighting the drivers of difference across it.
But that was just the start. Long-term goals, such as the leveraging of unsupervised machine learning and topographical analysis methods to deliver novel, powerful insights, meant that a scalable solution was necessary. And an experienced and knowledgeable partner like T-DAB.
In order to develop the data mining and dashboard applications to a production ready level (TRL9), the teams from T-DAB supported a knowledge-sharing programme and series of workshops. As well as identifying insights and specifying requirements, sessions like these are crucial to develop an understanding of the system’s infrastructure and architecture.
Working in close partnership, T-DAB conducted a data audit and suitability assessment of the platform’s existing data structure and format, assessing data quality for volume, velocity, veracity and variety, alongside whether it’s fit for purpose.
Exploratory Data Analysis offered insights from visualisation and ordination, whilst allowing for crucial data cleaning and wrangling before the introduction of bench-marking methods, including basic statistical modelling, visualization and plotting, funnel plots and outlier analysis.
The Solution
From the outset, the software provider stipulated that the project needed to enable its clients to carry out – and benefit – from analyses to benchmark manufacturing companies according to configurable KPIs.
To achieve this, T-DAB utilised a range of methods within the solution. Alongside statistical process controls, its propensity modelling offers insights into outliers.
Drawing inspiration from the analysis of ecosystems and populations, it analyses the manufacturing system in search of dissimilarities.
Data quality is crucial in a project like this, and with Cobot usage growing, it offers opportunities for data automation. The system’s architecture and front ends offer the cobot data capture and storage capabilities necessary to achieve this.
A tool like this is as powerful as its users need it to be, thanks to the development of an easy-to-use interface design hosted in the cloud. It also enables the software provider to cater to the needs of five different user groups. Quality Assurance users, for example, can access an analysis of data attribute usage, while Customer Process gives production monitors details of customer processing efficiency.
Management teams now have a clear view of operations, thanks to an intelligent oversight tool, which provides an analysis of employee performance based on the benchmark software tool data.
The Result
This project, which is just the start for the software provider, saw the organisation meet its aim by empowering its small and micro manufacturing client base with a first-class benchmarking insights they need to succeed.
Charged with making this vision a reality, T-DAB’s knowledge and guidance ensured the project ran smoothly, as well as successfully, maturing the data mining and analysis applications to production.
Best of all, T-DAB’s drawn on the latest generation of scalable and pervasive AI and machine-learning platforms to ensure an accessible solution for manufacturers, including small and micro, to benchmark their performance against, and developed a platform that can grow and evolve as its customers do.
About T-DAB.AI
T-DAB.AI is data science and data engineering innovation company. We develop innovative, bespoke machine learning-driven solutions to allow anyone to infuse technology with the spark of predictive intelligence.
