To help you gear up for the new year, National Instruments has published its NI Trend Watch 2016. In an introductory section, Eric Starkloff, executive vice president of global sales and marketing, says the insights in the report come from the vantage point of a company that invests more than 16% of its revenues in R&D. Topics include 5G prototyping, Big Analog Data, the Industrial IoT, smart-device test, and the consumerization of software.
I’ll be commenting on each of these topics, but will start with Big Analog Data—a topic I’ll also be addressing in a forthcoming print article on cloud computing. The challenges are daunting as we face a Cambrian explosion of data, with companies analyzing only 5% of the data they acquire.
According to the Trend Watch, the challenge can be met with smart enterprise management with edge analytics. It quotes —Dr. Tom Bradicich, general manager and vice president, Hyperscale Servers and IoT Systems, Hewlett Packard Enterprise, as saying, “Analytics at the edge in IoT and other industrial solutions play critical roles in solving the Big Analog Data problem. Intelligent measurement nodes afford analysis of data inline and, in turn, accelerate meaningful results.”
Intelligent data-acquisition are becoming more decentralized with processing elements moving closer to the sensor, the report notes, enabled by the latest silicon and IP from companies including ARM, Intel, and Xilinx. Consequently, smart software-based measurement nodes make it unnecessary to log every data point to disk.
When measured physical phenomena do occur that require human intervention, an enterprise data management system is necessary for getting the right data in front of the right people. That in turn requires properly documented data sets and smarter analysis.
Proper documentation begins with engineers agreeing on what metadata is important for analysis and defining metadata nomenclature and attributes. As for smarter analysis, the report cites Frost & Sullivan estimates that big-data analytics applied to testing can cut product development costs by 25%, operating costs by 20%, and maintenance costs by 50%. One goal is to find new correlations that can help predict future behaviors—key for maintaining a competitive edge.
View the entire Trend Watch at http://www.ni.com/trend-watch/.