The Coming Manufacturing CIO Automation Hangover

Thanks to COVID, manufacturing companies have been investing in automation at a record level over the last twelve months. Management consultancy McKinsey reports that in a survey of 800 executives 67% say they are increasing their automation investments either somewhat or significantly.  The business journal The Economist reports that automation investment nearly doubled in 2020 over 2019 primarily due to COVID.  As companies invest to reduce the density of employees on the shop floor, augment an aging workforce, or make up for the access to expert resources due to travel restrictions they are looking to smart systems, IoT, and automation as the answer to these challenges.  However, with all this investment in technology, CIOs are going to discover that this rapid digitization of the plant floor is going to create headaches when it comes to dealing with all the data that is being generated from these systems.  Imbibing too heavily in this newly data-rich environment is going to leave them with a workload hangover they are likely ill-prepared to deal with.  Once the data is out there, everyone is going to want access to it and unfortunately, much of the data isn’t all that easy to get to.

The Data Firehose

Since most smart systems and sensors can generate anywhere from several to dozens of data points and the spread of smart devices throughout manufacturing plants is growing at a rate between 50% and 100% in the last twelve months that means the data coming from operations is also growing exponentially.  The good news is that all this data means quality engineers can likely work through product issues from remote locations, be it their home or other locations where they don’t have to go to the plant.  The same is true for reliability and maintenance engineers or process optimization specialists.  The challenge is just getting them the right information from the right places at the right time.  It was tough enough when a plant might have a few thousand data points being generated at any given time.  When they suddenly have hundreds of thousands or millions of data elements the problem becomes much more complex.  For continuous process plants with data historians the problem has not been as bad since many already have historian technology that handles high volumes of time series data from multiple sources but with new devices often adding new non-time-series data to the mix the problem still is there.  For facilities without data historians, the problem is new and can threaten the viability of your digitization projects. This firehose of data can tax even the most capable IT department as they try to deal with all the various protocols and messaging standards for operational data that are in use today.  Ever since the 1980s, the challenge has been that 60% of the effort of integrating information has been in just getting the interfaces in place to move the data to where it is needed.  The same dynamic is now facing data scientists in building out analytics solutions to leverage this information.  In the last 40 years, the problem of getting data out of devices and using it to improve operations has changed only slightly.  Fortunately the manufacturing data integration problems today can be addressed to reduce the cost of getting and maintaining access to information from the plant floor.

Hangover Prevention Requires Preparation

Since the proliferation of IoT devices and the data explosion is happening, manufacturing CIOs can’t avoid the growth of data.  To avoid the assumption of massive technical debt in the form of interface development and maintenance they should take preventative measures to avoid the pending hangover.  The way to do that is to start now with preparing an industrial data architecture plan that puts tools in place that will help you make all the operational technology IIoT originated data digestible by your IT systems.  Whether these tools are brokers or applications which are capable of providing brokerage services will depend on your unique situation.  But failure to act now, as data volume is still expanding, will leave you with a hangover that could have been avoided.  Take the following steps to reduce the technical debt unconstrained automation could cause:

  • Extend your architectural efforts all the way down to the plant floor
  • Put in place a data ops strategy for operational data
  • Invest in data brokering solutions
Disclosure

The views and opinions in this analysis are my own and do not represent positions or opinions of The Analyst Syndicate. Read more on the Disclosure Policy.