The Technology Underpinnings of Geospatial Visualization
As previously published in CIO Journal from The Wall Street Journal:
Geospatial visualization is evolving into a powerful source of value. IT shops must get ready to tap its potential.
The human brain is naturally wired to process visual images by recognizing patterns, inferring relationships and discerning features. “Geospatial visualization” combines these perceptual and cognitive strengths with modern analytical computing capabilities that connect and combine data through geography. Building compelling geospatial visualizations from diverse information sources allows decision makers to gain compelling insights from tremendous volumes of complex and diverse data.
Is your IT shop ready to tap the potential of geospatial visualization? Full capability starts with tools and computational power for handling massive data volumes, rendering complex visuals and widely disseminating contextual geographic maps. The most advanced visualization and analytics tools are still useless without “good” data that is tested, correct and current. Reliable, scalable enterprise information management techniques and core informatics discipline are critical.
To help you understand the challenges, let’s take a closer look at how geospatial technology has evolved to become a powerful source of value and what it takes to unleash that power in your organization.
Ready for primetime
Geospatial capabilities themselves haven’t just arrived on the scene, although they were previously confined to use in specialized areas. The oil and gas industry has used geospatial analysis and mapping for decades in resource exploration and extraction analysis. Specific sectors of the federal government have used these tools to analyze population shifts, resource utilization, impacts of regulatory changes and more. What has changed to make both the availability and applicability of geospatial visualization so much greater today than in the past?
Early spatial modeling tools required proprietary knowledge and advanced training to create or import data, build explicitly geographic data models and create cartographically valid visualizations. Specialized resources and software were also required, limiting who could participate in discovery and analysis and how the results could be embedded into decision-making processes. Models were typically based on laboriously processed historical batch data disconnected from transactional systems, providing few opportunities for performance management and real-time operational reporting. What’s more, location-enabled data sets were scarce and expensive.
Today, more and more data already include geography, especially data from mobile, social and sensor-based sources. Ubiquitous GPS readings have vastly increased our ability to integrate data and geography. Techniques are also available to geocode legacy data containing addresses or place names, allowing companies to combine old with new data to analyze and visualize geographical trends and patterns over time.
Spatial tools and analytics have become much easier to use, apply and communicate. Advanced tools offer the ability to move beyond static, single-dimensional “push pin” views and interact visually with complex data. They allow correlation of diverse sources of multidimensional data, providing insight into trends and relationships over space and time. Today’s platforms allow complex data to be transformed into intuitive, interactive and descriptive views that are ready for exploration, discovery and decision-making by people who do not necessarily have specialized geospatial knowledge. And smartphones, tablets and other mobile devices have made these geospatial visualizations readily available almost everywhere that people work and make decisions.
Making the geospatial vision a reality
IT shops must be prepared to manage or exploit to unleash the power of geospatial visualization:
GPU-based processing. Systems based on graphic processing units (GPU) are used to apply techniques at the heart of geospatial modeling. These techniques include raster algebra across points, lines, or polygons; processing of high-resolution imagery collected from satellites and airplanes; and executing change detection of individual pixels (that is, determining if an image at one location or time differs from a separate image at a different location or time). Organizations are likely to need more computing power to apply these techniques. GPU-based systems can perform geospatial-modeling calculations over 100 times faster than general purpose CPUs.
Master data management. Relationships depend on identity, but linking distinct entities across enterprise systems and external feeds continues to be a critical challenge. Without correlation of master data, analyses could miss vital connections between data sets, possibly leading to bad decisions.
Data quality. Visualizations are only as good as the source data. Incomplete or incorrect location information compromises geospatial analysis. When external feeds from social media or other largely uncontrolled sources are incorporated, even organizations with broad data stewardship programs will have to deal with questions of dubious or unknown data quality. Completely cleansing potential inputs is not a viable option. Instead, organizations should tag questionable sources and try to utilize as many diverse data sources as feasible. In this way, end users can be aware of possible issues and have options for identifying and discounting unreliable data.
Big data and cloud computing. Geospatial visualization can allow organizations to tap into external sources of location-enabled unstructured data. This provides additional challenges for analyzing, storing and integrating “big data” into conventional business intelligence and warehousing solutions. Cloud-based storage and processing capabilities can give organizations a cost-effective way to scale geospatial visualization capabilities flexibly and provide them to their users as needed.
Augmented reality. Augmented reality (AR) systems connect information with reality. Some, for instance, use smartphone cameras and on-board processing to layer geospatial visualizations directly on top of real-world landscapes. Recent advances in indoor location-based services are likely to unlock new uses for AR throughout the enterprise, whether in retail stores, warehouses, or front-office headquarters.
Time and place underpin everything that happens in our lives and everything we know and learn about the world. Today’s technology allows us to collect information continually about nearly everything that affects us, fueling an explosion of real-time, location-enabled data. Organizations that successfully harness this data with the power of geospatial analysis and visualization stand to gain a significant competitive advantage.
Deloitte FAS LLP
Deloitte FAS LLP