Tuesday, April 3, 2012

How to Turn Oceans of Data into Manageable Streams #bigdata

This is a re-post from IBM's Building a Smarter Planet Blog.  Originally published there on April 3rd, 2012.

By Andrew Juarez
Lead Systems Specialist
Information Systems and Services
Coca-Cola Bottling Co. Consolidated

My company has a huge thirst for data. It’s one of the key resources that keeps our operations running smoothly as we produce roughly 150 million cases of soft drinks each year and distribute them to stores and restaurants throughout 11 southeastern states.

The products we sell aren’t high tech. They include sparkling beverages such as Coke and Sprite, bottled water, juices and sports drinks. But our business increasingly relies on cutting-edge technology to keep the bottling plants, warehouses and delivery trucks operating optimally in an increasingly competitive business environment. The pressures on us: We need to run lean and to get the freshest beverages to consumers when they’re thirsty.

Even as a lifelong IT guy, I’m continually surprised and thrilled to see the impact computers and software have on our business. In our 46 warehouses, for instance, forklift drivers have view displays that show them where every palette of products should be positioned on the floor so trucks can be loaded most efficiently. Our truck drivers and field technicians receive order and delivery  information updates during the day via mobile devices. We even inventory the spare parts on the vans our soda fountain equipment technicians drive. We consider them to be mini warehouses.

All of these handoffs of information require a sophisticated data management system. We need to mine the data quickly and continuously for business insights, so we have to make sure that the sheer volume of information doesn’t slow down the computer system’s response time. When we upgraded our system four years ago and started using IBM’s database software, the technology made it possible for us to reduce our data storage footprint by about 40%. At the same time, the growth rate of stored data was chopped in half.

Today, we’re a beta tester for IBM’s newest database software. Scientists from IBM Research have come up with new algorithms that make it possible to strip a lot of duplication out of the system. We expect this technology to compress the data by another 20%–saving us money and improving the performance of the entire enterprise.

In this era of big data, IT guys like me need help turning oceans of information into manageable streams.