Exelis Visual Information Solutions (VIS) just launched the ENVI Services Engine (ESE), an offering that takes desktop functionality and extends that to either a cloud or enterprise framework. Sensors & Systems (S&S) editor Matt Ball spoke with Beau Legeer, VP of Product Marketing at Exelis VIS, about the background on the development, the functionality and flexibility, and the future of the tool.
Legeer shared that the genesis of this development happened during a showcase of enterprise functionality with a prototype demonstration to the National Geospatial-Intelligence Agency (NGA) at the 2011 GEOINT Symposium. The exercise was to display desktop analytics available within ENVI, but putting those in an on-demand mode and applied to a browser or any device based on a specific problem rather than a piece of data on a system. The strong interest in the approach convinced Exelis VIS to productize that offering, with several customers deploying this approach and technology for their needs as the product has been in development.
The value from the application side of this approach is that users have all the advanced analytics and algorithms in ENVI available, they can also customize with IDL (the scientific programming language), and can provide custom applications in a cloud or enterprise environment through a web browser rather than a desktop environment. There are reference implementations provided as applications that include anomaly detection, pan sharpening, vegetation delineation, line of sight, and spectral identification to classify materials based on a hyperspectral dataset. With these capabilities, and the flexibility of the IDL programming language, users can deploy custom solutions to the multiple users and to the field.
The value from the sales side is that the approach can take the analytics you currently deploy and simplify the delivery, replacing the need to install software on every desktop and greatly increasing flexibility and speed of updates. Moving to the back-end enterprise also offers greater control and security of the data.
With the big analysis programs in the Intel space there can easily be 1,000 or more desktops, which translates into thousands of man hours saved by removing desktop maintenance. The solution can also scale down to a work group, where teams are building toolsets, collaborating and consuming imagery and services.
Anything that can run a web browser can run the custom applications and solutions. The clients are lightweight and don’t require a heavy client to do feature-rich mapping or to power the services engine. Combining OGC and Esri-based image standards and resources such as tiling and streaming, the solutions work to push services to clients even with low bandwidth. The technology is a simple service that allows any piece of middleware or client to access the computational resources powered by ENVI. With a web connected application you power a server-based algorithm by making a call over the web.
The enterprise integration is very straightforward, and doesn’t require any specific middleware. It works well if an OGC or Esri server is in the mix. If using Esri’s technology to build clients for iOS or Android, you leverage the ENVI ESE, and put the results directly over a map for context. The app-based approach strings tools together in a way where the end user doesn’t see or need to know the steps running in the background.
Exelis VIS creates custom clients for customers in their professional services group and have had great success in geospatial exploitation markets. One area of application has been a target detection workflow with hyperspectral imagery. Before, such a solution was in the realm of the image scientists where they were interpreting imagery and providing the detection results. Now, there is a workflow and solution that analyzes available imagery quickly, solving real-world problems with imagery in a “forward-facing” environment and moving more toward real-time.
With all the sensors available now, many of the solutions work to bring together multiple feeds. There can be still imagery taken a few minutes prior coupled with full motion video (FMV) from a UAV platform and a solution that pairs those images and allows the user to compare with change detection algorithms to point out what has changed. Parent company ITT Exelis brings much of this capability together in their Jagwire information processing and dissemination solution.
While ENVI Services Engine just came out, there is ongoing work to continue to improve the functionality and performance. Among the planned enhancements are new development tools to deploy services from the desktop with minimal coding, taking the tools that are put together in an analytical workflow and make those available as services. There are also plans to take the current ENVI for ArcGIS Server, and transform it to be powered by the underlying ESE where processing and client creation is done in the ArcGIS environment with their APIs for Web or mobile, and again with very little programming to deploy to the enterprise.
Exelis VIS is also taking some of the lessons learned with this enterprise offering to enhance ENVI for ArcGIS Server and power that with ESE, making it more scalable, multi-node, and high performant as a true server-class system with multiple cores or a cloud with multiple nodes where before it has run on a single server.
The cloud approach pulls image-based solutions away from heavy desktop and single-user frameworks and into very flexible and high performance offerings that can serve users wherever they are and with whatever device they have on hand. This focus of marrying imagery with GIS is squarely on solving problems, and exploiting imagery from any platform to efficiently address issues even in real time.