Building the Right Environment to Support AI, Machine Learning and Deep Learning
By Mark Kirstein, Senior Director of Enterprise Software, Zebra Technologies.
As new Internet of Things devices get churned out at accelerated rates, developers have adapted quickly by becoming proficient at creating mobile apps to connect consumers to their newly connected gadgets. Products such as the Nest home thermostat, the Staples Connect home automaton hub, the FitBit activity tracker, and the apps that help control them are popular examples of the IoT's growing influence in the consumer space. However, demand for apps that utilize the advantages of the IoT is also steadily rising in the enterprise, and this year we will see a greater number of developers embrace the more complicated challenges of creating IoT apps for enterprise needs. Why is app development for enterprise IoT more complex than for consumer IoT? For this reason: Whereas consumer IoT is often a 1:1 experience involving one person and one sensor, the enterprise requires more demanding IoT apps that often must work with hundreds of thousands of sensors.
Connected devices have remarkable potential for enterprise applications. Think of countless sensors along a pipeline, able to feed data into a system. IoT enterprise apps could send alerts to personnel in the field utilizing information from those sensors to drive real-time behavior and decision-making, assist in maintenance, reduce dangerous accidents in critical situations, and so forth. The apps will necessarily be robust and nuanced in dealing with vast amounts of sensitive and complex data. Add to this the app coding and the knowledge necessary for interfacing with specialized IoT sensors; it becomes clear that IoT apps for the enterprise require another layer of sophistication on top of what is needed for their consumer-class cousins.
One major enterprise IoT technology is RFID for location and motion tracking of objects, from storage containers to boxes to individual items. Combining RFID tags with enterprise IoT apps can achieve the full digitalization of all information about those items: the what, where, status, history, condition, and all other pertinent data for each individual unit. This leads to pretty meaningful advantages in the realms of procurement, manufacturing, and logistics, and offers the ability to manage it all—no matter the scale—from convenient app interfaces.
Imagine, for example, this technology applied to a grocery store, with workers handling 40,000 SKUs with 40,000 product labels, typical numbers for a chain store. Developers could seek to achieve new efficiencies in pricing changes by incorporating electronic shelf labels and Bluetooth low-energy technology to let employees make pricing changes dynamically. They might also use BLE to offer customized coupons to customers, all overseen from a store management app.
For another example, take a utility company with field technicians who need to interact with thousands of pressure sensors throughout a pipeline. The developers of the app used by those field technicians will certainly need to rely on features unique to the enterprise side of the IoT.
IoT enterprise app developers must be sure to capitalize on the major advantage of the technology: the ability for users to operate on real-time data and react to changing information. IoT apps offer unprecedented insights into current business activity, setting the stage for real-time operational intelligence, real-time actions and collaboration, and actionable knowledge that drives real-time decisions. Delivering all of this, along with interfaces that users can comprehend, will require a deep savvy with the technology, and clever design that makes those systems leveraging countless sensors and inputs informative and natural to use.
New interfaces and improved artificial intelligence may be the answer in helping employees to command these new tools, like the adaptive natural language understanding of api.ai for voice interfaces, or apps for the hands-free interfaces of wearables like Google Glass, or the newly announced Microsoft HoloLens. With products like FitBit in the mainstream, wearables in the IoT are gaining mindshare that is driving new ideas and new thought leadership that will push major growth in enterprise adoption. Next-level interfaces must be on app developers' minds as they work to build industry-specific technology applications that enhance worker efficiency.
In customer-facing arenas such as retail floors or the hospitality and healthcare industries, managers grapple with questions of how to equip every employee to provide great customer experiences. With IoT apps, developers will give them the tools to manage the whole of their operation like a chessboard, deploying employees to aid customers intelligently based on knowledge gleaned from all previous such interactions. For example, if data tracked by the IoT sensors on a retail floor discovers that customers tend to pick up a certain item and read the box but then leave without buying—but that customers do buy when the item is explained to them—customer service representatives could be alerted via app and deployed whenever that scenario is in play.
Similarly, enterprise IoT will streamline the movement of people, inventory, and assets, and bring about new levels of operational productivity. The degree to which these vast possibilities of enterprise IoT see successful implementation will depend upon the ability and execution of app developers in delivering on IoT's unique advantages.