S5 Manufacturing Analytics

Uplift and analyze manufacturing data

The Suite5 Manufacturing Analytics Suite is an enterprise data analytics engine that allows any manufacturer to uplift and analyze their legacy and historical data on a batch mode, in order to better understand certain aspects of their everyday operations (e.g. at the shopfloor for predictive maintenance purposes or in decision tables for augmented business intelligence purposes), and easily detect any emerging patterns and any unusual situations.

Manufacturers are sitting on a goldmine of unexplored historical, legacy and operational data from their Manufacturing Execution Systems (MES), Enterprise Resource Planning systems (ERP), Product Lifecycle Management systems (PLM), Manufacturing Resource Planning systems (MRP), Customer Relationship Management systems (CRM)… However, only 20-30% of the value from such available data-at-rest is currently accrued (Source: McKinsey, 2016).

The Suite5 Manufacturing Analytics Suite is addressed at different teams (managers and factory workers) who need to obtain a quick understanding of the data produced in their watch and gain insights on what happened in the past (for historical data) and what is happening at the moment or the near past/future (for operational data).

The main functionalities provided by the Suite5 Manufacturing Analytics Suite include:

  • Uplift, map and store datasets provided as csv files (and in the future through APIs).

  • Interactively visualize the data uploaded in order to easily and quickly slice it in different ways, check whether there are any missing or unexpected values and understand where it needs cleaning or further pre-processing.

  • Process and correlate the data provided per production line in order to provide insights in the current operations in relation to production at different granularity levels (per machine or per production line) in an intuitive manner.

  • Navigate to the outcomes of a set of machine learning algorithms that are currently supported:

    1. Time Series Prediction: LSTM (Long Short Term Memory) neural network

    2. Dimensionality Reduction: Kohonen’s Self-Organizing Maps (SOMs)

    3. Clustering: K-means

    4. Classification: SVM, Naïve Bayes, k-NN, Decision Tree, Random Forest, MLP

    5. Regression: Linear Model (Huber method), SVM (3 kernels: linear, RBF, Polynomial), Decision Tree

  • Sharing datasets at team level and different user access levels.

The Suite5 Manufacturing Analytics Suite is offered as a cloud service and practically provides an intuitive user interface for different business users based on the analysis configured by a data analyst from our team. Particular emphasis has been laid on explaining (to the extent that it is possible) the outcomes of the analytics in a user friendly and clear manner to the business users.

With a number of machine learning algorithms readily available in the Suite5 Manufacturing Analytics Suite and some more functionalities on the works, Suite5 always strives to further experiment with new manufacturing data and additional variations of algorithms to solve new problems in the manufacturer’s shopfloor and the decision table.