In the manufacturing world, a lot of lessons can be learnt, and insight gained by defragmenting data across a business.
Manufacturing businesses, even new ones, very rapidly find themselves beset by issues generated by fragmented data. What is more, many of them do not appreciate that it is happening to them. After all, their object in life is to design, make, or refine `X’, and their issues associated with that task are predominantly physical, or at least they appear to be. Yes, there is digital data involved, but it can often seem to be a side issue of little immediate consequence.
Start-ups in manufacturing would seem to have the best chances to deal with issues of data fragmentation, but that does still pre-suppose that they are aware of the problem. There will undoubtedly be many advances in manufacturing technology that they can take on board and exploit. Not least the ability to take a computer-generated design and feed it straight into a 3D printer system and have it immediately realised in a saleable form. They will also be using the latest design tools, and well as online cloud services for running much of their day-to-day business management activity.
But one of the troubles with managing day-to-day activities is that the mind span of the business can easily slip into that 5-day, 8-hours a day cycle when it comes to managing the data that’s generated. Life becomes about just ensuring that the right amounts of the right raw materials and pieces are available from suppliers when they are needed. It becomes about knowing what needs to be produced, in what quantity, where it has to be delivered, and when. It is also, of course, about the most critical day-to-day job of managing the money: which customers have paid and which haven’t, which suppliers need paying and which might be tolerant.
But that daily grind lets one crucial aspect of modern business escape if business managers are not careful: what can they learn from all that near- and longer-term data that has been generated. There is, nearly always, much to learn – and a good deal to earn – from being able to examine that historical data. The one trouble then can be finding it and then knowing which parts are valid and which are in some way spurious. With fragmented data in different silos around the business, it is all too easy to generate invalid results. The data is misunderstood or misinterpreted as versions of the same data points are being stored in different data siloes generated by various applications.
This problem usually is many times worse for well-established manufacturing businesses, for the undeniable reasons that they have a good deal more historical data to deal with, and that much of it is now in siloes generated by legacy applications that are either not used frequently, or are even no longer used at all. That means the data may be hard to locate, or only available in formats that are not compatible with any of the applications currently in use. Yet the information that is held can be of immense value right across the business, especially where the lessons of days gone by can educate a new management team from making a mistake.
Finding ways to defragment the organisation’s data is a vital capability for all manufacturing businesses, not least because the volume of data they are collecting continues to grow. The need now, as competitive pressures continue to grow, is to expose and exploit the value that is held within it.
The first job is to find that data. Some estimates suggest that as much as 80 percent of it is effectively hidden away in the applications that do not form the frontline, mission-critical applications and processes running the day-to-day business. Much is held in what has become known as secondary storage – backups of historical production data archived away for safe keep. And therein lies one of the critical problems for businesses, in that even calling such data `secondary’ or ‘dark’ hides that fact that, these days, all business data can have a role on the front line.
For just one example, data held in test and development archives may show that early attempts with a possible new manufacturing process might fail for any number of different reasons. But combining that data with later information on new materials, tools or process techniques could show that the failed manufacturing process would now work profitably.
But to find that out means having the ability to locate and access that data and then analyse it in conjunction with data from other sources – from both within the business and from other external sources – to expose any new opportunity. Perhaps the most critical factor is being able to access the data in the first place so it can be made available throughout the business. The key, therefore, is the ability to locate and access data that is now spread around a range of different storage services – physical on-premise storage, cloud and even tape archives.
Once located, the task of extracting new value from the data can begin. This requires manufacturing businesses to have the right tools in data management and analytics to make that happen. Still, there are now tools that will work across on-premises, cloud and virtual storage environments.
Taking advantage of these tools brings together the data location, security, management and analysis in a single environment, rather than businesses having to create yet another silo of `the data we have sent for analysis’.
Regardless of whether the companies are small, large, new, or indeed, well established, being able to bring together and ask questions of all the information and data stored within the business is going to drive new efficiencies in manufacturing and this could very well be the competitive advantage that businesses need right now.