In the final article of this three-part series (see Part 1 and Part 2), we'll look to the horizon and beyond for some of the key forces shaping the future world of location intelligence. We'll focus on a few that have already begun to appear and are likely to evolve rapidly over the next few years.
Big Data and the Internet of Things
The rapid explosion of data from deep within web logs, back-office systems and the billions of location-aware devices we all use every day has created new land for location intelligence to thrive. Today the physical world is being linked to the Internet as an active participant in generating even more data. Smart meters are a good example. They have a location and generate information on electricity usage that gets communicated to a backend ERP
or marketing system that, in turns, triggers a billing process or generates a message to entice you to change your behavior, like reducing consumption during peak hours.
Web 2.0 was all about user (human) generated content; the Internet of Things extends that to inanimate real-world objects. Like humans who interact with business processes through customer portals, mobile apps, call centers and other touch points, objects in the Internet of Things also interact with business processes and each other by providing information or taking some action based on a rule or algorithm (such as turning off your power if your bill is 90 days past due!).
Real-world objects like smart meters have been part of the landscape for some time. SCADA (supervisory control and data acquisition) systems have long monitored widely distributed infrastructure for power generation and distribution, water and petroleum pipelines, ports and airports, transportation systems and more. Fixed and mobile sensors are widely deployed to monitor the air, water and ground for everything from environmental health to earthquake warning systems. People have long been transporting sensors with them in the form of cell phones and tablets. Recently, wearable sensors have started to appear for the mass market such as Google Glass or brainwave reading headbands from InteraXon
. Companies such as Under Armour
are developing sensor clothing today for NFL players, tomorrow for all of us? Estimote
are taking advantage of new Bluetooth low energy technology to build small, low-cost sensors for fixed and mobile deployment. Whether you're in a building like a retail store or you've misplaced your purse, these small sensors will bring you location-relevant information to help you shop or find misplaced things.
Although limited in what these devices sense, all that is changing rapidly. Sensors will soon pack in more and more capabilities to sense temperature, light, sound, position, acceleration, vibration, stress, weight, pressure, humidity and other environmental measures into smaller and smaller packages, some no bigger than a grain of rice. They will be carried by us, worn by us, stuck on walls, embedded in things and scattered around, providing an unending stream of data for yet unimagined applications. According to ABI Research, some 30 million wearable health devices alone were shipped in 2012; and by 2020, 30 billion wireless devices of all kinds will connect to the "Internet of everything," sixty percent of them sensor type devices.
Taming Big Data
Location Intelligence will play a very critical role in organizing and using the tsunami of data generated from this Internet of everything. Bringing Big Data Together
Location is a common key for bringing disparate data together whether from database or application silos, smartphones or tablets, or sensors of all kinds. Data related to a common location lets you merge or "mash up" data from any source with a spatial reference of some kind. The spatial relationship doesn't have to be perfectly coincident but "close enough" to establish a relationship that provides new insight and enables deeper analysis. For example, if your GPS says your car is in the middle of a lake, there's a good chance that it's really on a nearby road; and based on proximity, location intelligence tools can "snap" the car and all of its onboard sensor information to the closest road for use in profiling driving patterns. This can likewise be applied to Tweets, Facebook postings or any other data that has some explicit (x, y or grid coordinate) or implicit (address or place name) location identified. The ability to relate and bring together seemingly unrelated data makes location intelligence an indispensable tool for taming and using big data in business intelligence
(BI) and enterprise applications. Seeing Patterns and Trends
Once brought together, location intelligence applications facilitate data exploration through visualization and analytic tools. Simply seeing data in a map view is often enough to gain new insight, and more can be teased out with spatial analysis tools such as heat maps or spatial statistics that can help predict where populations of customers will move over time. Visualization is already a proven tool for taming big data. Adding spatial visualization and analytics amplifies its value. Finding Needles in the Haystack
OMG TMI! Too much information! The phrase applies to big data as it does to information on our personal lives or others. Big data is personal and non-personal data brought together with the hope of making sense of it and driving some type of action. Location and spatial relationships provide a powerful filter for selecting relevant data from the haystack. I have lots of information about customers who have visited my thousands of coffee shops, but I'm really only interested in the ones within walking distance, for example. Spatial relationships also can play a key role in taking action. If a customer gets within walking distance of my coffee shop, send him a message to stop by for the holiday latte.
As the Internet of Things generates unfathomable amounts of data, the ability to use spatial relationships to filter data relevant to where I am right now, at this time, and to trigger user-defined events such as turning on the heat when I'm 30 minutes from home will play an crucial role in finding what's relevant and useful in both our personal lives and at work within our BI, ERP, CRM and other tools. Without these spatial filters and geotriggers, our interaction with the Internet of Things would be severely limited by TMI!
At least three key technology trends are driving how we will access and use location intelligence in the future: the evolution has already begun. Cloud Computing
According to David Linthicum of InfoWorld, "Cloud adoption's tipping point has arrived." Though somewhat lagging this tipping point today, the world of location intelligence is also experiencing a global warming trend toward cloud computing. Well established in the cloud are commodity assets such as base maps (streets, topographic, etc.) and satellite imagery from a number of sources. Nearly every BI, ERP, CRM, SCM
and other business system embeds cloud-based maps for a limited number of applications. When the inflection point is soon reached for the location intelligence cloud, it will trigger a massive migration of internal spatial data to the cloud which will, in turn, pour down thousands of small, focused web apps for embedding in our work apps and personal apps alike. Why can't I see my business travel and personal travel on the same map? Why can't my geotrigger let my colleagues know I'll be 30 minutes late for the meeting and, at the same time, let my wife know I'll be 30 minutes late for dinner? The line between the office and the home becomes obscured in the cloud. Mobility
The first set of mobile waves that washed over the consumer market are being followed by another set poised to flood the business world with consumer-like applications for the field that will change the paradigm for how we think about BI
and other business systems. Desktops and servers sitting back at corporate will no longer be the center of attention but one of several systems enabling the efficiency and effectiveness of both white- and blue-collar field workers. Everything from how apps are designed to where data is captured, stored and used will center around the mobile user and the devices they use to generate sales, provide service and gather field information for future strategic and tactical moves.
Location intelligence will play an obvious key role in organizing this data, using it to optimize visitations, locate parts, capture competitive intelligence, sense the environment and facilitate communication with other mobile users. District managers armed with the latest reports on sales and profitability will pull in competitor locations, predicted weather, drive times and other location data to create an order for next week’s delivery. Sales staff will be pushed real-time opportunities or complaints near them. Service workers will be routed to meet at some location to exchange needed parts. Maps take center stage as the user interface for the field. As Noam Barden, CEO of Waze (now a Google company) puts it, "You find your way around the Internet with a search bar; to find your way around the real world you start with a map." In-Memory Computing
While mobility is all about getting data to the field, in-memory computing is about bringing data and analytics to the CPU. Historically, location intelligence apps have required that data be extracted from the analytic platform, such as a data warehouse, and brought into the spatial analysis tool for processing. This follows a historically similar pattern for other analytic tools such as SAS or SPSS. Today, there is a growing capability and trend to leave the data in the warehouse and bring the analytics to it, to be done in-memory without moving data and intermediate results to disk and back. For example, SAS runs analytics in the Teradata database today, and in the future will run them in SAP
HANA's in-memory platform.
As these in-database and in-memory platforms increase their core foundational spatial capabilities, location intelligence vendors will follow suit, building applications optimized for an in-memory platform that enables entirely new spatial analytics and workflows that were impractical or unthinkable using the old extract/disk ways. On the horizon you can see a new generation of location intelligence apps built from the ground up for in-memory spatial analytics. Embedded in BI tools and other business applications, this will improve both real-time predictive models and the actions prescribed by them.
Embedded Location Intelligence
Location intelligence embedded in BI and other business applications is itself another major trend that will continue. The first generation of embedded location intelligence is here. Nearly every BI and enterprise application vendor has static or dynamic maps based on standard geographies to link with charts and graphs within the application itself. The future will bring much more powerful visualization, the ability to use custom boundaries, like your sales territories, not just standard geographies such as ZIP code, county or state. Spatial data from outside the report, application or the organization will be accessible on demand within the BI tool or enterprise application to provide additional context for decision making, such as real-time weather or demographic data, facilitating better decisions and operational BI.
Beyond visualization, query based on spatial relationships and bringing in outside contextual spatial data, BI and enterprise apps of the future will include advanced location analytics for a host of use cases such as site selection, field optimization and demand forecasting. These may stand alone or work in concert with other analytic tools such as SAS or SPSS as part of a larger model, bringing in new data sets, enhancing algorithms with spatial mathematics, improving the results and helping communicate them to managers via simple, intuitive maps.
In this last piece of our three-part series, we've looked at a few of the common forces that are driving the future of location intelligence along the same paths that BI and enterprise apps are following. We've avoided science fiction and focused on what is actually in play today and likely to be here and become even more important tomorrow. Most importantly, throughout the entire series we've tried to relate the world of location intelligence to the experiences and language of the BI community with a little history, a description of its landscape and identification of the parallel paths we follow. A major goal was to remove any mystery, fear, uncertainly or doubt about bringing these paths together in your organization. I hope we've been successful!
SOURCE: The World of Location Intelligence Tomorrow
Recent articles by Steve Benner