There was a very interesting article from
Dan Kusnetzky in ZDNET about Big Data. Telco has always been at the forefront of producing, processing, and analyzing large amount data. What I'm witnessing today is another order of magnitude of data volume. When items such as razor blades, racks of fresh produce, jeans, and pallets are "sending" updates about their locations, status, and temperature through a network of automated sensors, we are talking "big data". That data tells a lot about habits and trends at a micro (privacy) and macro (economy) scale.
If you were to send all this raw data right to a central server, you would make Big Data simply Huge Data, and immediately lose the value of real-time automated data. We believe a large amount of information generated at the Edge should be used at the Edge. Only a fraction of that data should make it to a central server and become Big Data.
When a forklift picks up a pallet full of fresh berries from a loading dock, then drives through a multitude of portals equipped with sensors to finally put away the fruits in a cold room, thousands of records have been created. They include very important real-time information such as: "Is this the right truck and loading dock?", "Is this the right pallet?", "Are these the right berries?", "How many workers are around the forklift while it is backing out?", "Are all the pallets in the right cold room?". This data is critical during the actual movement of the assets but unless exception occurs, we are only interested in a small subset for the long term.
I've been a little perplexed by some of the recent HP moves, but this one makes a lot of sense to me. Any large IT vendor has to provide a complete set of integrated IT management, automation and security software to its customers. The datasets generated by enterprise applications are growing so large that software built by companies like Autonomy, Vertica (acquired earlier this year by HP as well), Greenplum (acquired by EMC) is becoming mission critical.The market for unstructured data is getting bigger, and Oracle is the only one that hasn't made a move yet.
CHEP is a global leader in managed, returnable, and reusable packaging solutions, serving many of the world's largest companies in sectors such as consumer goods, fresh produce, beverage, and automotive. CHEP’s service is environmentally sustainable and increases efficiency for customers while reducing operating risks and product damage.
Omnitrol Networks has been named one of the Top 5 vendors in manufacturing intelligence by Manufacturing Automation readers. This is the result of Omnitrol offering a complete pre-integrated solution that provides predictive analytics based on information collected by sensors. The challenge is not only to manage the vast amount of data coming in real time from a multitude of automated sensors but to make sense of what needs to happen immediately versus what needs to be archived for further data mining.