IoT: Game Changer for Data Science
Growth in IoT will put pressure on organizations to step up their data science so they can develop and deploy analytics that tap the new data streams.
- By David Stodder
- February 3, 2016
As if the world were not already swimming in data, news from the annual Consumer Electronics Association (CES) trade show in Las Vegas in January made it clear that much more is on the way. CES has become a showcase for the Internet of Things (IoT). Consumer products such as cameras, virtual reality headsets, vehicles, and drones are bristling with sensors that can collect data and share it with analytic applications and operational data systems over networks. IoT growth will require more organizations to develop a strategy for how they plan to manage new flows of big data into operational systems as well as for the data science and analytics necessary to derive business value.
Vehicles, for example, are becoming nothing less than big data on wheels. The self-driving cars that garnered so much media attention at this year’s CES are equipped with sensors that feed data to both embedded and remote software systems that monitor and analyze every facet of the vehicle’s performance and behavior. Nvidia’s DRIVE PX2 computing system, along with other software tools, modules, and artificial intelligence, gives vehicles “360-degree situational awareness” according to the company. Vehicles with telematics and other sensor-dependent systems will generate new data relationships involving location, movement, and proximity. Analysis of these relationships will impact nearly every field, including insurance, advertising, marketing, logistics, and resource planning.
Once people step out of their vehicles, smart “wearables” embedded with IoT will communicate data about health or the presence of environmental hazards that might trigger allergic reactions, including location and other information that could be useful to marketers and advertisers. Sports apparel companies, such as Under Armour, are advancing the “quantified self” trend by creating wearables that can analyze data generated by exercising. Professional athletes, many of whom are already using analytics to improve performance, will likely pioneer advanced uses of wearables and influence younger athletes who emulate them.
Wearables will also be important for professionals in construction, manufacturing, transportation, energy, shipping, and other industries where it is important to monitor alertness and other behavior. Working alongside humans will be robots that will also generate streams of data flowing into operational and analytics systems.
IoT and Data Science
Growth in IoT will put pressure on organizations to step up their data science so they can develop and deploy analytics that tap the new data streams. Of course, this is easier said than done because there remains a shortage of data scientists, with those at the top of the field proving to be “unicorns” who are hard to find and even harder to keep. In the mania for data science, organizations must choose applicants carefully so they do not land “fake” data scientists who lack the skills and experience to deliver value.
Data science is about applying scientific methods to explore and test hypotheses about the data. It is a term that unites contributions from several fields, including statistics, mathematics, operations research, computer science, data mining, machine learning (algorithms that can learn from data), software programming, and data visualization. It can cover the entire process of acquiring and cleaning data, methods for exploring the data and extracting value from it, and techniques for making insights actionable for humans and automated processes. Most often, the focus of data science is to optimize decisions and realize higher value from data through advanced analysis.
Data scientists are highly varied in their experience because few, if any, people possess the background and skills in all of these fields and processes. Some are R or Python programmers; others are data visualization experts. Willingly or not, some focus most of their time on data preparation steps and are thus expert in creating good data sets for analysis and modeling. Others excel at a more conceptual level and can learn the business domain and articulate how business decisions, processes, or customer engagement could benefit from data science.
For this reason, the best practice is to create data science teams rather than depend on one or a few individuals. Project teams can bring together business and technical personnel; many may already be part of the organization. Team leaders should have good communication skills so they can explain project objectives and results to business leadership. Teams will need access to data visualization tools to improve speed to insight and make it easier to explore data and share analysis.
IoT could be a game changer for many organizations, or perhaps better put, IoT will be a game changer for innovators who figure out how to disrupt industries through fast, innovative analysis of new IoT data sources. Organizations that have been sitting on the sidelines regarding data science from afar need to begin developing a strategy or risk being left behind.
David Stodder is director of TDWI Research for business intelligence. He focuses on providing research-based insight and best practices for organizations implementing BI, analytics, performance management, data discovery, data visualization, and related technologies and methods. He is the author of TDWI Best Practices Reports on mobile BI and customer analytics in the age of social media, as well as TDWI Checklist Reports on data discovery and information management. He has chaired TDWI conferences on BI agility and big data analytics. Stodder has provided thought leadership on BI, information management, and IT management for over two decades. He has served as vice president and research director with Ventana Research, and he was the founding chief editor of Intelligent Enterprise, where he served as editorial director for nine years.