From the Editor's desk: Data data everywhere, but the value is in analytics
September 2017, News
As the Industrial Internet of Things gains momentum, the sheer quantity of data it brings becomes difficult to comprehend. For instance, it is estimated that one ‘autonomous’ car will consume around 4 terabytes of data every day. Extrapolate that to a future ‘digital’ factory, which may contain thousands of sensors, and you quickly realise there is a tsunami on the horizon.
The point of it all is to capitalise on the actionable insights rumoured to lie buried somewhere deep within those terabytes. To use a mining analogy, the data are the tons of bedrock that must be processed before a single ounce of gold can be recovered. The data then is only of any practical value after it has been processed (mined) by the tools needed to extract the information that lies within it, and the purpose, in an IIoT context, is to drive new value creation through access to information that was not previously available to managers or other relevant plant personnel.
The tools in this case are the analytic packages capable of crunching through data that was not optimally utilised before. This is an area where many companies struggle at present. While they have the expertise and process knowledge necessary to set up the sensor networks and other data-capturing infrastructure, they lack the ability to develop and code the analytical algorithms that are necessary to give it the context required before value-adding actionable insights emerge in areas like predictive maintenance, efficient energy utilisation, supply chain optimisation, or to initiate a timely response before a critical event, results in lost production, or worse.
New opportunities for vendors
There has been a flurry of recent activity from many vendors in this regard, aimed at claiming a share of what is touted to become a $500 billion a year market by 2020.
Some of the big automation vendors are positioning their offerings as integration platforms for IIoT analytics. Rockwell Automation is one of them with its web-based FactoryTalk business intelligence solution that integrates all data into one information management and decision support system. The company recently extended the platform through the launch of the FactoryTalk Analytics for Devices appliance and the FactoryTalk TeamONE app. The idea is to enable teams on the plant floor to make quicker decisions via the information generated by the appliance and then distributed on the app. These off-the-shelf offerings require minimal configuration and are designed to solve common maintenance problems faster. Together, they provide an aid to improve the reaction time of maintenance teams and assist managers through equipment health and diagnostic analytics.
These are some of the first subscription offerings from Rockwell Automation, accessible via a new self-service e-commerce portal. (See ‘Getting started with analytics’).
Endress+Hauser has taken a different approach. The instrument company has gathered a great deal of deep domain expertise in a wide variety of industries since its founding nearly sixty-five years ago. The water industry is a good example, and it comes as no surprise that E+H is currently involved in a project with the Federal Ministry of Economic Affairs and Energy in Germany to help develop an intelligent measuring system for the water supply of tomorrow.
Process data from a number of multi-parameter Endress+Hauser measuring stations is transmitted to a central server at the Technical University of Cologne where it is analysed using a series of proprietary mathematical models. The objective is to identify those parameter sets that can be used to detect the optimal operating point of a wide area water supply system. These can then be used to provide the country’s utilities with new opportunities to improve the efficiency and stability of their own delivery systems. (See ‘From data to insight’).
There is no doubt that we are about to enter an era that will generate more data than we can possibly imagine. Collecting that data will be easy, but the organisations most likely to thrive on it will be the ones that best understand their specific business problems and how analytical models can be applied to create new competitive advantages. Perhaps the fourth industrial revolution would be more aptly named as ‘The age of analytics’.
Editor: SA Instrumentation & Control