Great article explaining big data using the three Vs — volume, velocity and variety. I liked it because it explained Big Data in simple terms.
Following a period of disillusionment over the unrealized promise of big data, companies are starting to make it work, says Gary Survis, a Wharton lecturer and senior fellow at Wharton’s Initiative for Global Environmental Leadership. At the same time, since 90% of all data ever created has appeared in just the last two years, big data is also breaking the systems that hold and measure it. From seeds that can divulge how well they are growing to jet engines that collect megadata for fuel saving, some returns are coming in. In this Knowledge@Wharton video interview, Survis discusses just what constitutes big data and what it means for business today.
Some memorable quotes:
Personally I think we may not be able to use the same old information to solve new problems. The existing sensors we have in plants today were put into plants to solve control problems. They do a great job of that. However, they may not help us solve reliability, maintenance, energy efficiency, and HS&E problems. That is, just historizing data from existing sensors may not be the solution to solve these new challenges. I think you need to take a look at what problems you have, then figure out what information you need to solve those problems, and then deploy software and raw data sensors accordingly - if the data isn’t already being collected. So we must ask what the problems are:
Is it a Reliability problem? – Are equipment like pumps, fans/blower, conveyors, agitators, and motors etc. failing without warning causing plant downtime? If so, deploy missing wireless vibration and temperature sensors on these equipment together with asset monitoring software which is able to detect signs of trouble early so you can fix it quickly before a long outage results.
Is it a Maintenance problem? – Is equipment like heat exchangers, air cooled exchangers, and cooling towers being taken offline and torn apart only to find they are not fouling and are not damaged – causing downtime, long turnarounds, high cost, and distracting resources away from equipment that really need attention? If so, deploy missing wireless pressure, temperature, and other sensors on these equipment together with asset monitoring software which is able to tell if equipment needs cleaning or not.
Is it an Energy problem? – Is the energy cost high but nobody knows exactly where the energy in the form of fuel, steam, compressed air, and chilled water etc. goes, or if it is wasted? If so, deploy missing wireless flow transmitters together with energy management software for unit-wise energy metering to detect inefficiencies and for cost accounting. Deploy missing acoustic transmitters on steam traps together with steam trap monitoring software to detect where failed steam traps are blowing steam. Fouled heat exchangers or un-optimized cooling tower fans are detected by the asset monitoring software.
Is it an HS&E problem? – Are bypass valves left open after proof testing? Do spills happen because the position of other manual valves is not known? Is response to activation of safety showers and eyewash stations taking too long? Is the timing and release from relief valves not known etc.? Are operators exposed unnecessarily collecting data manually by reading gauges, sight glasses, and VA meters, maybe using dip sticks in the field? How about technicians using portable vibration testers and temperature guns? Various missing wireless transmitters for position, vibration, pressure, temperature, pH, conductivity, level, and flow etc. can be deployed to collect that data and integrate it into the DCS and historian.
So I personally agree data is important to generate the information you need to make decisions. Start by analyzing what problems you have and what data you need. If you already have the data, fine. If not, deploy sensors for missing measurements as necessary. In an existing plant, wireless sensors are the easiest to deploy. Most of the data will not go the DCS because it will not be used by the operators, it will go to the historian because it is used beyond the control room by other disciplines.
Get more insight from the articles below:
Second Layer of Automation
www.ceasiamag.com/.../10354
Maintenance with a Hart
www.ceasiamag.com/.../9894
Smart Sensing: Situational Awareness
www.iaasiaonline.com/more.php
Are you sending people in harm's way?
Wireless for Asset Uptime
www.ceasiamag.com/.../8689