Over the past few decades, manufacturing operations have undergone a major evolution from being passive and largely human controlled to extremely complex and intelligent. One major driver in this evolution has been the way manufacturing data is handled, analysed and used in making real time manufacturing decisions.
According to Manufacturing and the Data Conundrum, a study conducted by the Economist Intelligence Unit regarding data collection and its use in manufacturing, the amount of shop-floor data being collected has increased manifold for a majority of the respondents. This is largely from the widespread availability of real time sensor data at various stages of the manufacturing process which is essential to integrated operations. But, on the other hand, only a minority of manufacturers have advanced data analytics in place to predict issues. Fewer still have managed to integrate data from diverse sources or have skilled personnel to analyse it.
Through effective and integrated data management and analysis from various manufacturing process, a significant impact can be made to various facets of manufacturing – from product quality management and process controls to safety and facilities management. The study mentions examples of companies like Meritor, ABB and GE, where increased focus on quality through integrated systems was demanded by new contracts and proves an often debated point on the economics of investing in data analytics for product quality management.
Implementing an integrated and effective data collection and analysis architecture does come with its share of hurdles. Many manufacturing facilities are still on old legacy systems which generate different forms of data. Managing the data in a coherent and common format can be a gargantuan task and moving to a common architecture is not without its birthing pains. Big data also presents with it the problem of staffing the infrastructure with trained personnel. Most often than not, different manufacturing facilities and even departments within the same facility are silos in operation which resist any change to status quo. Rallying them on the importance of moving to common data architecture and an integrated view of operations requires a strong leadership drive.
All said and done, the next leap in manufacturing process control is bound to come from effective and integrated management of big data. 34% of the surveyed companies report annual savings of more than 25% through effective implementation of sophisticated, integrated digital control systems. The survey also found that companies which have a well-defined data strategy are more likely to be profitable with average earnings growth of more than 10% annualized.
In conclusion, there is no denying the fact that advanced, integrated data capabilities are beneficial to the companies in the long run. Also, availability of data has never presented a problem with most modern manufacturing units being capable of churning out terabytes of data from thousands of sensors. The challenge lies in the capability to effectively store, manage and analyse integrated data to give actionable insights.
What are your views on integrated manufacturing operations? Please feel free to write in.