The integrated sensor and systems approach was on full display at Hexagon’s recent conference that combined the technologies and audiences of Intergraph/ERDAS, Leica Geosystems, and the metrology division. The event provided the opportunity for editor Matt Ball to speak with Mark Doherty, CTO of Intergraph’s Security, Government and Infrastructure Division about the ongoing develop of integrated solutions, touching on various sensor feeds, and the ongoing move to incorporate more real-time inputs.
Ball: To start, John Graham, SG&I’s president, mentioned that 62 products were released last year in the division. That’s a huge number. What is involved in meeting that aggressive timeline?
Doherty: It is a big effort. Part of that is that we have an industry focus, and not just a platform, so as soon as you move up that stack into industries, you invariably have to build solutions and products that fit those business needs. The downside of that is that it leads to a proliferation of products. But, if you’re going to be in those industry spaces, you must have a solution. It’s a challenge to keep that machine moving forward with all those releases, but we’ve learned to do that over the years.
Having said that, we’re always looking for ways to consolidate and keep the portfolio manageable, because that’s in our interest and in our customer’s interest. Keeping things tight is better in terms of the effort that we put into each individual product. In that 62 number, there are a range of releases, with some new products, significant feature upgrades, and more minor upgrades with patches, platform upgrades, and fixes. It’s still a release that requires testing, certification, and effort to get it out the door.
Ball: In terms of the operating system environment, I know you’ve added Android and iOS on the mobile side, and that with each operating system it requires whole new products and releases.
Doherty: We’re just cutting our teeth on iOS and Android right now. On the development side, there are some pretty good tools out there with HTML5 and PhoneGap and those technologies, that let you maintain a single code base to create an app that is targeted to different platforms. But, particularly in the public safety space, we still need to do a rigorous testing on those platforms, because it has to work.
It’s easy to create a new piece of software in some sense, and I think a lot about this in the mobile space, because you can sit down with a team of one or two guys and they can knock something out in a week. But, in our enterprise spaces we’re making a commitment to our customer that we’re going to support that for some time. You know, if Angry Birds goes away, we’ll get another Angry Birds. If we release something, regardless of what mobile operating system, they’re expecting that as part of their enterprise solutions that it will be around and supported. You can’t enter into those releases and commitments lightly, because the application halflife is still going to be there for some time to come. It’s different than the things that you find on the App Store, because if it vanishes off your phone, it probably wouldn’t impact you much.
Ball: There are so many conservative organizations out there that wait and wait to upgrade computer hardware and software. The have and have nots really come to the forefront when you talk about some of the big wins in such places as India, where paper-based systems are now being replaced. It’s still amazing that such a legacy still exists.
Doherty: Inertia is a wonderful thing, isn’t it? You used India as an example, but I guarantee you that it’s not limited to the developing world. We could walk into many agencies in North America or Europe, and find workflows that are paper based or that have had their first round of process re-engineering 15 years ago. If you now compare that to what you can do with Web-based and mobile technologies, it could be much more efficient.
Back to the enterprise, once you build those processes that touch a lot of things and that are integrated with a lot of systems, they get a certain amount of inertia. You don’t change those things on a whim on a Tuesday afternoon, it’s a process to cut that over to something new.
Many of our big projects, such as utility or public safety, the technology is a big part, but they’re also big change management projects. You’re having to deal with new processes and people using new tools that are different than what they’ve been using for 12 or so years, and that is a big undertaking.
Ball: The move to mobile is probably one of the starkest changes there in terms of giving up a degree of control on the individual worker and where work gets done, but with the portability so much can be gained.
Doherty: I think we have a good advantage, because mobile is not new to us. In our utility and public safety space, we’ve had tightly integrated mobile solutions for years. Those were running on more traditional form factors, such as ruggedized laptops mounted in the vehicle. The good part about that legacy is that we understand the enterprise part of that, with the round-tripping of workflows, and being able to operate in a sometimes-connected environment, and not having applications that just don’t function. We have a lot of experience there.
Now we’re taking the mobility experience and bringing it to new form factors -- an iPhone or Android device -- but still with that understanding of what is required to integrate it properly in enterprise workflows.
Ball: In your presentation this week, security seemed to be a big focus, with a lot of innovation there. Is that where you spend a good deal of your time in product development?
Doherty: My time gets sliced across all of our portfolios, but security is certainly a growth area. We’re in a transition with security and monitoring, and showing that those kind of applications aren’t just security. We did show an example of a bad guy coming into a secured area, but certainly with our integration with Leica, those same technologies are applicable to the monitoring of landslides or a construction site. We see that as a growth area, that range of technologies to take sensor data and bring it into some kind of decision and visualization engine.
We had a good investment in that area before Hexagon bought us, but I’d say that the Hexagon acquisition has reinforced that and has accelerated some things in that area. Here on the show floor we are featuring a dam water monitoring project, where we integrated a number of sensors, some from Leica and some from third parties, into a visualization and monitoring environment for a complete turn-key solution for the full range of dam issues. Certainly, we here at Hexagon see that as a huge growth area.
Ball: That ties nicely to what Informed Infrastructure is about, where monitoring and adapting to change is increasingly important. The change management solutions almost beg a services approach, such as in the dam monitoring realm where the full system provides a service. Would you sell a combination of sensor hardware and software as a service model?
Doherty: I think there is definitely a potential for that. I’ve been talking with a number of people lately, and now we’re segueing into the whole cloud computing area. I’m sure you’re familiar with the Gartner hype cycle. I think in the cloud space, we’re trending down into the trough of disillusionment, although that sounds harsher than it is in this case. We’re getting to the point where everyone is sort of looking back, and understanding where it makes sense to apply this technology.
In general, we’re looking for opportunities where we can enter into that Software as a service (SaaS) model where it makes sense. In some of our markets, such as security, we’re being approached by some smaller operators who are asking to buy our software as a service, but the vast majority of them wouldn’t be comfortable without having all that functionality inside their own firewall. There is also a practical side to public safety, where if that infrastructure fails, and I lose my Internet connectivity, then I have no service. In other areas, such as the monitoring example, it might make good sense as as service.
I was recently speaking with Bo Peterson, Hexagon’s CTO, and you hear this number of 87,000 dams in China. In that, and around the world, there is a range from really small facilities to really large. To penetrate the lower end of that market, being able to offer a service, makes sense. Selling the hardware, but having it tie into cloud computing resources with a web front end for monitoring, would be a great way to penetrate those smaller systems. It’s certainly something that we’re looking at in a lot of places, to understand where it makes sense to provide a hosted solution, but the market conditions have to be just right to make customers want to do that.
Ball: With decreased budgets, where capital costs aren’t as likely of an investment, a service might be a really attractive investment.
Doherty: Moving to services is a double-edged sword for software and hardware vendors, but history tells us that you ignore those trends at your own peril. If you decide not to try new models, because it affects software sales and maintenance revenue, then you face an innovator’s dilemma from those that take that approach and price you out of the market. So, you have to go at it with your eyes open and take some steps to be able to compete.
The biggest place that I’m seeing that right now is with the U.S. federal government, that are aggressively moving to hosted solutions and cloud-based offerings. I think they’re quite open about it as a means to reduce cost of infrastructure and finding ways to avoid up-front capital expenditure and move it toward a pay-as-you-go model. I think when Obama came in, his IT team had this as a fundamental transformation idea.
Ball: Along those lines, a very recent development is the idea that Data.gov won’t be a repository for data, but a means of open access through APIs like the FCC’s Broadband Map. That’s fascinating, because the role of government changes somewhat.
Doherty: It’s funny, because in spatial data infrastructure (SDI) context we really feel that we were doing cloud before it was cloud. You look at cloud closely and it’s what SDI is about, it’s services that are running somewhere that you really don’t know about, and to some extent don’t care about as long as you can access the data in a way that is useful. That’s what cloud is.
I think we shouldn’t lose sight of that SDI vision. Some of our competitors are asking for all of your data, kind of the opposite of what SDI is supposed to be. We’re taking a different stance. The piece that’s missing for me with SDI is that front end. We’ve been through the cataloging, but we still need a place where a casual user can easily get access to the data sets without having to be an expert user and knowing how to find all this stuff. If we come up with better ways to deal with that, then all the infrastructure is there and hosted by federal governments. There are tremendous examples of data portals that are there to be used, certainly in Europe with the INSPIRE initiative.
Ball: On the sensor side, you certainly have a lot experience with video sensors. Is there a lot of work on development around the video feed?
Doherty: We started about six or seven years ago on that whole security area. We had an initial contract in the New York area with the Transit Authority, where as part of the contract we got into the management of video. There are different elements, with an increasingly-intelligent camera, some local storage, and a video management server. Where we plug-in is with the video management server.
We had to learn a lot about interfacing with that server, and to effectively stream that video with the bandwidth that is available. That is what is built into our iSight product, that manages the load so that you don’t bring servers to their knees. There are also little things, such as dealing with camera contention, where if you have five or six operators working on an incident, you have to manage who has control in a given scenario. It doesn’t do anyone any good if we’re all trying to pan a camera simultaneously, and we end up locking it up. Those are some of the practical things that we’ve come a long way with over the past seven years.
Moving beyond security, we’re seeing more interest in police having integrated video in their own environment. Not from a Big Brother is watching scenario, but from an incident management tool to help you make better decisions. Taking what we’ve learned in the security space, and applying it to a policing case. All those cameras are out there to be used, and I think our customers are asking how to integrate them and bring them into their operational environments.
Ball: When it’s real-time it’s fairly intuitive to add the visual piece, but is there analysis too, where it gets fed into a reporting system or other analysis functions?
Doherty: We do a little bit of that. The biggest investment we’ve made there is with our Motion Video Analyst software, that operates in a real-time mode to take a stream off a UAV or other georeferenced video source to show that video over terrain in real time. It also operates in a forensic mode where you can query a video repository and access those videos through metadata and the locations, and browse that after the fact.
The Holy Grail there is doing in-video analytics, such as “find me the blue car.” I know that is a difficult nut to crack. We’re generating these massive amounts of data, but if you have to manually sift through it all, it’s a huge task to make use of it. If there are ways you can throw intelligent queries at video data, and get a list of frames of interest, it can really streamline things.
Ball: I’ve heard in that security space that we’re just a few years away from algorithms that will predict human behavior.
Doherty: And what does that mean, right? We’ve shown quite a bit with smart cameras that are pretty good at detecting which human left a bag where or after tracking an object it suddenly becomes two objects and alerts to that problem. Some of that we have figured out, but going to the next step with predictive behavior is interesting.
Ball: The capture and display of 3D is another area of activity, where you have visualization capability in GeoMedia and also hardware in the Leica Geosystems area where you capture high-resolution models. Are these two areas coming together?
Doherty: GeoMedia 3D is out there, and it lets you do a great job of assembling multiple data sources into an integrated model, and doing some analysis with that. 3D is definitely an area where our acquisition by Hexagon has ratcheted up the emphasis and the urgency for us to do more in that 3D space.
As you say, now as a Hexagon family we have that complete workflow of sensor data capture of 3D data in a number of forms (whether top-down Lidar or ground-based laser scanning) and then pushing it through a complete workflow. As part of a Hexagon entity, we want to find ways to create a competitive advantage by offering something that the industry thinks is unique and groundbreaking.
I think we’re at a bit of a crossroads with 3D, we can all visualize and zoom around models, but to me the next step is to do something transformational, but not so hard to use that it is too technical for people. How do you get 3D integrated into workflows and software as a natural way that people work.
You probably saw a bit more about the Hexagon acquisition of myVR, that has tools for publishing data to the web and mobile, and with the ability to manipulate and do some analysis on those models. I think that you’ll see more of that from us in the not too distant future, with a focus on those areas.
If you step back and look at Intergraph as a whole, in our PPM and SG&I divisions, if the holy grail is to bring building information modeling with GIS, there should be no company that is better positioned to do that than Intergraph. We have detailed BIM in our PPM division, with a sophisticated engineering-grade model. We’re increasingly developing the tools in our SG&I division on the geospatial side of that. Figuring out ways to meld those two together would seem to be something we should figure out how to do. I think Hexagon is nudging us in the direction to figure that out.
Ball: In all the vertical markets where you play, you’ve set the foundation for a lot of the additional advancements, but then it gets back to services. It’s got to be tough on the main stage, because there have been a lot of platform advancements, but what’s sexy is the solutions side. Telling the platform story isn’t as interesting, it’s how the platform is applied.
Doherty: We’re doing a bit of both, we’ve put a renewed focus on products such as GeoMedia and ERDAS, and selling those very capable geospatial tools as products. But that’s a thing that is different about Intergraph is that we’re a solutions company with a public safety, security, water and electric utility solution. Our customers have a business problem to solve, and what we try to do is instead of saying here are a bunch of tools, here’s a solution set that solves the business problem for you.
Ball: To close, I’d like to delve into the real-time data space some more. Are you doing more to integrate real-time data across systems?
Doherty: Absolutely. We made an acquisition last year of Augusta Systems that has a product called EdgeFrontier. It is a sophisticated middleware that lets you quickly stand up interfaces and messaging, with inbound and outbound, and a rules engine in the middle. We’re very rapidly adopting that as a real-time hub and message switch that takes incoming feeds (whether SensorML or XML), applies some business logic to it, and then decides what to do with that.
We’ve heard Hexagon CEO Ola Rollén’s vision of a dynamic GIS, with sensors informing the model, and that’s really what it’s about, taking more and more data, that’s more real time, and being able to update that model to reflect what the situation is on the ground more quickly. The EdgeFrontier product is a big part of that, having that engine as a part of our integrated offering to do that kind of processing.