Client Results

Industry 4.0 Project
Tait Electronics

Overview

Product Similarity Modelling

Tait are one of NZ’s largest electronic goods manufacturers.  They utilise a number of robotics lines to build the Surface Mount Technology (SMT) boards that make up their product range.  The short-run size nature of the Tait product line and market means that there are large efficiency gains available if product similarities can be understood and this learning is applied to optimising the robotic line set up.

Activity

This project grouped surface mount technology (SMT) products such that the overall component mix within each group was optimised for a single robotic SMT line. This resulted in fewer changeovers and therefore greater productivity.  The main tools and approaches being utilised from the I4 suite are:

  • High volume data capture and time-series tracking
  • IoT connectivity of systems
  • Machine Learning to divide the entire product base into homogenous groupings

The first phase of this project is complete, Tait is now using the groupings determined by the machine learning to optimise their robotics production lines.

 

Contact us to get started now

Complete the questions below to test your data maturity.

Over the next two years, which three of the 14 key performance indicators do you most want to improve on as a business?

Make a note of these before you carry on reading.

The key 14 performance indicator categories:

Productivity

  • Asset & equipment efficiency
  • Inventory efficiency
  • Materials efficiency
  • Utilities efficiency
  • Workforce efficiency

Flexibility

  • Planning & scheduling effectiveness
  • Production flexibility
  • Workforce flexibility

Speed

  • Time to market
  • Time to delivery

Quality

  • Product quality
  • Process quality
  • Safety
  • Security

Now ask yourself – what is your current performance against these three KPIs? Can you tell me how you performed in the last hour, yesterday or last week?

If you can’t answer this question for all three because you aren’t measuring the data, then the next step is clear. Figure out what data you need to enable you to measure it, and decide how you are going to collect that data.

If you can answer it historically; last week or last month – ask yourself, is this retrospective view sufficient for me to really make improvements?

If you can answer it for all three up to the minute, then it is quite possible that shopfloor intelligence isn’t a number one priority for you. Look out for parts 2 and 3 of this blog series for some more insights into how you can make the data work for you.