At one point or another, you may have come across the buzzphrase – Big Data. What is it, why does it matter and how can instrumentation users position themselves to ride this trending phenomenon?
Big data is a term used to describe the tremendous volume of structured and unstructured data that a business receives. The ‘Big Data Universe’ is growing exponentially. We are producing and consuming data at an unprecedented rate. Computer Sciences Corporation predicts that data production will be 44 times greater in 2020 than it was in 2009 with individuals creating 70% of all data and enterprises storing and managing 80% of it.
It is worthwhile to note that big data is more than just volume (e.g. petabytes or exabytes). In addition to volume, SAS Institute lists 4 other aspects (velocity, variety, variability and complexity) associated with big data. Here is a simple example of how big data is contextualized in an industrial plant.
- Volume – Big data in a process plant originates at the sensor level. There is tremendous amount of process information being captured across the plant’s labyrinth of sensors.
- Velocity – Data that is fed into the control system (e.g. SCADA) should stream in at near-real time speeds and dealt with in a timely manner. Any latency in transmissions could pose a danger.
- Variety – Data can come in from various sources and formats such as electronic data recorders, PLCs, HMIs and data historians (Dan Hebert, Four ways to Collect Process Plant Data).
- Variability – Data flow can be inconsistent with periodic peaks (e.g. Peaking power plants where there could be daily, seasonal or event-triggered peak data loads).
- Complexity – With a complex myriad of different types of information generated in a plant, it is necessary to connect the data dots and establish information hierarchies for effective plant monitoring and control
Data is only as useful as what you do with it and what you can make sense of it. In the words of Thomas Davenport an analytics thought-leader, “The sweeping changes in big data technologies and management approaches need to be accompanied by similarly dramatic shifts in how data supports decisions and product/service innovation”. Big data when processed, analyzed and interpreted correctly is significant because past business performance can be evaluated and future performance can be quantitatively predicted. This arms decision makers with descriptive, diagnostic, predictive and prescriptive information to shape the best corporate strategy.
As an instrumentation user, here are some ways (adapted from Littlefield, M (2015, November/December).Big data analytics. InTech, 12-15) to prepare yourself as your organization jumps on the Big Data bandwagon.
- Stay open and up to date with technological changes. According to InTech’s magazine cover story on big data analytics, “as legacy and next generation application build on top of IIoT platforms, new applications will be held to a higher standard than ever before in the industrial space” (Littlefield, M (2015, November/December).Big data analytics. InTech, 12-15) . Big data vendors need to create not only navigation friendly applications (think consumer apps like Uber) but also ones that weave social, collaborative, search tools and intuitive analytics into its architecture. Imagine, monitoring the state of your power plant’s instrumentation on a mobile platform interface. To some, that concept might be revolutionary while to others, it might be a precarious thought to have something so important hinged on an interface that appears to be skeletal. This will require a paradigm shift.
- Always refer back to the corporate strategy to maximise the value of big data analytics. Are you measuring the right process variables at the right places at the right frequency? Can the measurement data you are currently collecting be repurposed to indicate another aspect of operational or maintenance status? Can you expand the capability of your instrumentation to obtain additional measurement points? Also, as more companies adopt big data analytics, you can expect more cross functional collaboration between various departments (e.g. EHS, quality, asset performance management) in your organization to bring more structure to corporate goals (Littlefield, M (2015, November/December).Big data analytics. InTech, 12-15). If you are in a position to influence, be the voice that keeps everyone on track and gather only information that matters.
- Skills upgrade. Companies that embark on big data programs will invest in training their employees with the hope of ‘home-growing’ data science teams. In the long run, this is more cost effective from an organizational standpoint instead of outsourcing data services. Actively participate in these trainings, volunteer if need be, or even consider investing in a big data analytics course that is relevant to your area of expertise. Knowledge is power.