Adhering to standard business processes can sometimes cost a lot of money. Commercial flights, for instance, are mandated by law to adhere to a strict schedule of maintenance and inspection. The loss to airline companies due to these routine checks can be as high as $10,000 for each hour of non-operation. A recent study conducted by SAP showed that by using big data and in-memory technology to analyse airline engine data any by streamlining this maintenance process, unplanned downtime dropped by as much as 18%.
So what exactly does streamlining entail? In the case of the airline industry, operators set up sensors to measure more than 300,000 different parameters relating to the engine of the flight. This amounted to nearly 20 terabytes of information for each hour of the flight. Analysing these different parameters against benchmarked indices was sufficient to accurately predict any possible malfunctions. Doing such big data analysis not only prevents unplanned downtime, but also helps operators optimize their flight runs to account for future maintenance work, reduce the time it takes to inspect the symptoms (since the exact engine parameters are known) and thus avoid flight delays.
Such data driven streamlining of business processes is today possible in every industry; be it managing prospecting and lead generation in sales, or using data gathered from clinical surveillance platforms to minimize the spread of Hospital Acquired Infections (HAI). Big data has been playing a key role in streamlining processes that often hinged on guess work and approximations not too long ago.
Besides the obvious improvement in accuracy, there are other benefits to using big data to streamline business processes. One of the key upsides is in the subsequent automation of labour. IT teams routinely spend inordinate amount of time testing software systems to identify defects. With big data, it is possible to at least partly replace such human labour with automated cloud-testing tools. This saves time and also increases productivity – both of which play a key role in improving profits as well.
But one of the key challenges here is in identifying the actual parameters to be studied and designing the big data workflow accordingly. While big data systems are intended to process millions of data points to arrive at the output, it can also be quite resource intensive. As a result, measuring even a couple of parameters that have no direct impact on the output can exponentially escalate the resources required to handle the process. At the outset, this may seem like a trivial issue to handle. But when this is not done right, the costs accrued due to the addition of even one or two non-core parameters could escalate the overall budget, thus denying the true benefits of big data driven streamlining to the organization.
From a COO’s perspective, using data input tools like sensors can be a lucrative gold mine to measure the operational efficiency of a business with the help of big data and thus streamline the information. But equally important is the need to identify parameters that are critical and those that are not. Failing to do so could mean the difference between success and failure of the big data implementation.
