A discussion on how IT management technologies and methods have evolved to optimize and automate workloads to exacting performances and cost requirements.
Many of the latest technologies -- such as Internet of Things (IoT) platforms, big data analytics, and cloud computing -- are making data-driven and efficiency-focused digital transformation more powerful. But exploiting these advances to improve municipal services for cities and urban government agencies face unique obstacles. Challenges range from a lack of common data sharing frameworks, to immature governance over multi-agency projects, to the need to find investment funding amid tight public sector budgets.
The good news is that architectural framework methods, extended enterprise knowledge sharing, and common specifying and purchasing approaches have solved many similar issues in other domains.
BriefingsDirect recently sat down with a panel to explore how The Open Group is ambitiously seeking to improve the impact of smart cities initiatives by implementing what works organizationally among the most complex projects.
The panel consists of Dr. Chris Harding, Chief Executive Officer atLacibus; Dr. Pallab Saha, Chief Architect at The Open Group; Don Brancato, Chief Strategy Architect at Boeing; Don Sunderland, Deputy Commissioner, Data Management and Integration, New York City Department of IT and Telecommunications, and Dr. Anders Lisdorf, Enterprise Architect for Data Services for the City of New York. The discussion is moderated by Dana Gardner, Principal Analyst at Interarbor Solutions.
Here are some excerpts:
Gardner: Chris, why are urban and regional government projects different from other complex digital transformation initiatives?
Harding: Municipal projects have both differences and similarities compared with corporate enterprise projects. The most fundamental difference is in the motivation. If you are in a commercial enterprise, your bottom line motivation is money, to make a profit and a return on investment for the shareholders. If you are in a municipality, your chief driving force should be the good of the citizens -- and money is just a means to achieving that end.
This is bound to affect the ways one approaches problems and solves problems. A lot of the underlying issues are the same as corporate enterprises face.
Bottom-up blueprint approach
Brancato: Within big companies we expect that the chief executive officer (CEO) leads from the top of a hierarchy that looks like a triangle. This CEO can do a cause-and-effect analysis by looking at instrumentation, global markets, drivers, and so on to affect strategy. And what an organization will do is then top-down.
In a city, often it’s the voters, the masses of people, who empower the leaders. And the triangle goes upside down. The flat part of the triangle is now on the top. This is where the voters are. And so it’s not simply making the city a mirror of our big corporations. We have to deliver value differently.
There are three levels to that. One is instrumentation, so installing sensors and delivering data. Second is data crunching, the ability to turn the data into meaningful information. And lastly, urban informatics that tie back to the voters, who then keep the leaders in power. We have to observe these in order to understand the smart city.
Saha: Two things make smart city projects more complex. First, typically large countries have multilevel governments. One at the federal level, another at a provincial or state level, and then city-level government, too.
This creates complexity because cities have to align to the state they belong to, and also to the national level. Digital transformation initiatives and architecture-led initiatives need to help.
Secondly, in many countries around the world, cities are typically headed by mayors who have merely ceremonial positions. They have very little authority in how the city runs, because the city may belong to a state and the state might have a chief minister or a premier, for example. And at the national level, you could have a president or a prime minster. This overall governance hierarchy needs to be factored when smart city projects are undertaken.
These two factors bring in complexity and differentiation in how smart city projects are planned and implemented.
Sunderland: I agree with everything that’s been said so far. In the particular case of New York City -- and with a lot of cities in the US -- cities are fairly autonomous. They aren’t bound to the states. They have an opportunity to go in the direction they set.
The problem is, of course, the idea of long-term planning in a political context. Corporations can choose to create multiyear plans and depend on the scale of the products they procure. But within cities, there is a forced changeover of management every few years. Sometimes it’s difficult to implement a meaningful long-term approach. So, they have to be more reactive.
Create demand to drive demand
Driving greater continuity can nonetheless come by creating ongoing demand around the services that smart cities produce. Under [former New York City mayor] Michael Bloomberg, for example, when he launched 311 and nyc.gov, he had a basic philosophy which was, you should implement change that can’t be undone.
If you do something like offer people the ability to reduce 10,000 [city access] phone numbers to three digits, that’s going to be hard to reverse. And the same thing is true if you offer a simple URL, where citizens can go to begin the process of facilitating whatever city services they need.
In like-fashion, you have to come up with a killer app with which you habituate the residents. They then drive demand for further services on the basis of it. But trying to plan delivery of services in the abstract -- without somehow having demand developed by the user base -- is pretty difficult.
By definition, cities and governments have a captive audience. They don’t have to pander to learn their demands. But whereas the private sector goes out of business if they don’t respond to the demands of their client base, that’s not the case in the public sector.
The public sector has to focus on providing products and tools that generate demand, and keep it growing in order to create the political impetus to deliver yet more demand.
Gardner: Anders, it sounds like there is a chicken and an egg here. You want a killer app that draws attention and makes more people call for services. But you have to put in the infrastructure and data frameworks to create that killer app. How does one overcome that chicken-and-egg relationship between required technical resources and highly visible applications?
Lisdorf: The biggest challenge, especially when working in governments, is you don’t have one place to go. You have several different agencies with different agendas and separate preferences for how they like their data and how they like to share it.
This is a challenge for any Enterprise Architecture (EA) because you can’t work from the top-down, you can’t specify your architecture roadmap. You have to pick the ways that it’s convenient to do a project that fit into your larger picture, and so on.
It’s very different working in an enterprise and putting all these data structures in place than in a city government, especially in New York City.
Gardner: Dr. Harding, how can we move past that chicken and egg tension? What needs to change for increasing the capability for technology to be used to its potential early in smart cities initiatives?
Framework for a common foundation
Harding: As Anders brought up, there are lots of different parts of city government responsible for implementing IT systems. They are acting independently and autonomously -- and I suspect that this is actually a problem that cities share with corporate enterprises.
Very large corporate enterprises may have central functions, but often that is small in comparison with the large divisions that it has to coordinate with. Those divisions often act with autonomy. In both cases, the challenge is that you have a set of independent governance domains -- and they need to share data. What’s needed is some kind of framework to allow data sharing to happen.
This framework has to be at two levels. It has to be at a policy level -- and that is going to vary from city to city or from enterprise to enterprise. It also has to be at a technical level. There should be a supporting technical framework that helps the enterprises, or the cities, achieve data sharing between their independent governance domains.
Gardner: Dr. Saha, do you agree that a common data framework approach is a necessary step to improve things?
Saha: Yes, definitely. Having common data standards across different agencies and having a framework to support that interoperability between agencies is a first step. But as Dr. Anders mentioned, it’s not easy to get agencies to collaborate with one another or share data. This is not a technical problem. Obviously, as Chris was saying, we need policy-level integration both vertically and horizontally across different agencies.
Some cities set up urban labs as a proof of concept. You can make assessment on how the demand and supply are aligned.
One way I have seen that work in cities is they set up urban labs. If the city architect thinks they are important for citizens, those services are launched as a proof of concept (POC) in these urban labs. You can then make an assessment on whether the demand and supply are aligned.
Obviously, it is a chicken-and-egg problem. We need to go beyond frameworks and policies to get to where citizens can try out certain services. When I use the word “services” I am looking at integrated services across different agencies or service providers.
The fundamental principle here for the citizens of the city is that there is no wrong door, he or she can approach any department or any agency of the city and get a service. The citizen, in my view, is approaching the city as a singular authority -- not a specific agency or department of the city.
Gardner: Don Brancato, if citizens in their private lives can, at an e-commerce cloud, order almost anything and have it show up in two days, there might be higher expectations for better city services.
Is that a way for us to get to improvement in smart cities, that people start calling for city and municipal services to be on par with what they can do in the private sector?
Public- and private-sector parity
Brancato: You are exactly right, Dana. That’s what’s driven the do it yourself (DIY) movement. If you use a cell phone at home, for example, you expect that you should be able to integrate that same cell phone in a secure way at work. And so that transitivity is expected. If I can go to Amazon and get a service, why can’t I go to my office or to the city and get a service?
This forms some of the tactical reasons for better using frameworks, to be able to deliver such value. A citizen is going to exercise their displeasure by their vote, or by moving to some other place, and is then no longer working or living there.
Traceability is also important. If I use some service, it’s then traceable to some city strategy, it’s traceable to some data that goes with it. So the traceability model, in its abstract form, is the idea that if I collect data it should trace back to some service. And it allows me to build a body of metrics that show continuously how services are getting better. Because data, after all, is the enablement of the city, and it proves that by demonstrating metrics that show that value.
So, in your e-commerce catalog idea, absolutely, citizens should be able to exercise the catalog. There should be data that shows its value, repeatability, and the reuse of that service for all the participants in the city.
Gardner: Don Sunderland, if citizens perceive a gap between what they can do in the private sector and public -- and if we know a common data framework is important -- why don’t we just legislate a common data framework? Why don’t we just put in place common approaches to IT?
Sunderland: There have been some fairly successful legislative actions vis-à-vis making data available and more common. The Open Data Law, which New York City passed back in 2012, is an excellent example. However, the ability to pass a law does not guarantee the ability to solve the problems to actually execute it.
In the case of the service levels you get on Amazon, that implies a uniformity not only of standards but oftentimes of [hyperscale] platform. And that just doesn’t exist [in the public sector]. In New York City, you have 100 different entities, 50 to 60 of them are agencies providing services. They have built vast legacy IT systems that don’t interoperate. It would take a massive investment to make them interoperate. You still have to have a strategy going forward.
The idea of adopting standards and frameworks is one approach. The idea is you will then grow from there. The idea of creating a law that tries to implement uniformity -- like an Amazon or Facebook can -- would be doomed to failure, because nobody could actually afford to implement it.
Since you can’t do top-down solutions -- even if you pass a law -- the other way is via bottom-up opportunities. Build standards and governance opportunistically around specific centers of interest that arise. You can identify city agencies that begin to understand that they need each other’s data to get their jobs done effectively in this new age. They can then build interconnectivity, governance, and standards from the bottom-up -- as opposed to the top-down.
Gardner: Dr. Harding, when other organizations are siloed, when we can’t force everyone into a common framework or platform, loosely coupled interoperability has come to the rescue. Usually that’s a standardized methodological approach to interoperability. So where are we in terms of gaining increased interoperability in any fashion? And is that part of what The Open Group hopes to accomplish?
Not something to legislate
Harding: It’s certainly part of what The Open Group hopes to accomplish. But Don was absolutely right. It’s not something that you can legislate. Top-down standards have not been very successful, whereas encouraging organic growth and building on opportunities have been successful.
The prime example is the Internet that we all love. It grew organically at a time when governments around the world were trying to legislate for a different technical solution; the Open Systems Interconnection (OSI) model for those that remember it. And that is a fairly common experience. They attempted to say, “Well, we know what the standard has to be. We will legislate, and everyone will do it this way.”
That often falls on its face. But to pick up on something that is demonstrably working and say, “Okay, well, let’s all do it like that,” can become a huge success, as indeed the Internet obviously has. And I hope that we can build on that in the sphere of data management.
It’s interesting that Tim Berners-Lee, who is the inventor of the World Wide Web, is now turning his attention to Solid, a personal online datastore, which may represent a solution or standardization in the data area that we need if we are going to have frameworks to help governments and cities organize.
A prime example is the Internet. It grew organically when governments were trying to legislate a solution. That often falls on its face. Better to pick up on something that is working in practice.
Gardner: Dr. Lisdorf, do you agree that the organic approach is the way to go, a thousand roof gardens, and then let the best fruit win the day?
Lisdorf: I think that is the only way to go because, as I said earlier, any top-down sort of way of controlling data initiatives in the city are bound to fail.
Gardner: Let’s look at the cost issues that impact smart cities initiatives. In the private sector, you can rely on an operating expenditure budget (OPEX) and also gain capital expenditures (CAPEX). But what is it about the funding process for governments and smart cities initiatives that can be an added challenge?
How to pay for IT?
Brancato: To echo what Dr. Harding suggested, cost and legacy will drive a funnel to our digital world and force us -- and the vendors -- into a world of interoperability and a common data approach.
Cost and legacy are what compete with transformation within the cities that we work with. What improves that is more interoperability and adoption of data standards. But Don Sunderland has some interesting thoughts on this.
Sunderland: One of the great educations you receive when you work in the public sector, after having worked in the private sector, is that the terms CAPEX and OPEX have quite different meanings in the public sector.
Governments, especially local governments, raise money through the sale of bonds. And within the local government context, CAPEX implies anything that can be funded through the sale of bonds. Usually there is specific legislation around what you are allowed to do with that bond. This is one of those places where we interact strongly with the state, which stipulates specific requirements around what that kind of money can be used for. Traditionally it was for things like building bridges, schools, and fixing highways. Technology infrastructure had been reflected in that, too.
What’s happened is that the CAPEX model has become less usable as we’ve moved to the cloud approach because capital expenditures disappear when you buy services, instead of licenses, on the data center servers that you procure and own.
This creates tension between the new cloud architectures, where most modern data architectures are moving to, and the traditional data center, server-centric licenses, which are more easily funded as capital expenditures.
The rules around CAPEX in the public sector have to evolve to embrace data as an easily identifiable asset [regardless of where it resides]. You can’t say it has no value when there are whole business models being built around the valuation of the data that’s being collected.
There is great hope for us being able to evolve. But for the time being, there is tension between creating the newer beneficial architectures and figuring out how to pay for them. And that comes down to paying for [cloud-based operating models] with bonds, which is politically volatile. What you pay for through operating expenses comes out of the taxes to the people, and that tax is extremely hard to come by and contentious.
So traditionally it’s been a lot easier to build new IT infrastructure and create new projects using capital assets rather than via ongoing expenses directly through taxes.
Gardner: If you can outsource the infrastructure and find a way to pay for it, why won’t municipalities just simply go with the cloud entirely?
Cities in the cloud, but services grounded
Saha: Across the world, many governments -- not just local governments but even state and central governments -- are moving to the cloud. But one thing we have to keep in mind is that at the city level, it is not necessary that all the services be provided by an agency of the city.
It could be a public/private partnership model where the city agency collaborates with a private party who provides part of the service or process. And therefore, the private party is funded, or allowed to raise money, in terms of only what part of service it provides.
Many cities are addressing the problem of funding by taking the ecosystem approach because many cities have realized it is not essential that all services be provided by a government entity. This is one way that cities are trying to address the constraint of limited funding.
Gardner: Dr. Lisdorf, in a city like New York, is a public cloud model a silver bullet, or is the devil in the details? Or is there a hybrid or private cloud model that should be considered?
Lisdorf: I don’t think it’s a silver bullet. It’s certainly convenient, but since this is new technology there are lot of things we need to clear up. This is a transition, and there are a lot of issues surrounding that.
One is the funding. The city still runs in a certain way, where you buy the IT infrastructure yourself. If it is to change, they must reprioritize the budgets to allow new types of funding for different initiatives. But you also have issues like the culture because it’s different working in a cloud environment. The way of thinking has to change. There is a cultural inertia in how you design and implement IT solutions that does not work in the cloud.
There is still the perception that the cloud is considered something dangerous or not safe. Another view is that the cloud is a lot safer in terms of having resilient solutions and the data is safe.
This is all a big thing to turn around. It’s not a simple silver bullet. For the foreseeable future, we will look at hybrid architectures, for sure. We will offload some use cases to the cloud, and we will gradually build on those successes to move more into the cloud.
Gardner: We’ve talked about the public sector digital transformation challenges, but let’s now look at what The Open Group brings to the table.
Dr. Saha, what can The Open Group do? Is it similar to past initiatives around TOGAFas an architectural framework? Or looking at DoDAF, in the defense sector, when they had similar problems, are there solutions there to learn from?
Smart city success strategies
Saha: At The Open Group, as part of the architecture forum, we recently set up a Government Enterprise Architecture Work Group. This working group may develop a reference architecture for smart cities. That would be essential to establish a standardization journey around smart cities.
One of the reasons smart city projects don’t succeed is because they are typically taken on as an IT initiative, which they are not. We all know that digital technology is an important element of smart cities, but it is also about bringing in policy-level intervention. It means having a framework, bringing cultural change, and enabling a change management across the whole ecosystem.
At The Open Group work group level, we would like to develop a reference architecture. At a more practical level, we would like to support that reference architecture with implementation use cases. We all agree that we are not going to look at a top-down approach; no city will have the resources or even the political will to do a top-down approach.
Given that we are looking at a bottom-up, or a middle-out, approach we need to identify use cases that are more relevant and successful for smart cities within the Government Enterprise Architecture Work Group. But this thinking will also evolve as the work group develops a reference architecture under a framework.
Gardner: Dr. Harding, how will work extend from other activities of The Open Group to smart cities initiatives?
Collective, crystal-clear standards
Harding: For many years, I was a staff member, but I left The Open Group staff at the end of last year. In terms of how The Open Group can contribute, it’s an excellent body for developing and understanding complex situations. It has participants from many vendors, as well as IT users, and from the academic side, too.
Such a mix of participants, backgrounds, and experience creates a great place to develop an understanding of what is needed and what is possible. As that understanding develops, it becomes possible to define standards. Personally, I see standardization as kind of a crystallization process in which something solid and structured appears from a liquid with no structure. I think that the key role The Open Group plays in this process is as a catalyst, and I think we can do that in this area, too.
Gardner: Don Brancato, same question; where do you see The Open Group initiatives benefitting a positive evolution for smart cities?
Brancato: Tactically, we have a data exchange model, the Open Data Element Framework that continues to grow within a number of IoT and industrial IoT patterns. That all ties together with an open platform, and into Enterprise Architecture in general, and specifically with models like DODAF, MODAF, and TOGAF.
Data catalogs provide proof of the activities of human systems, machines, and sensors to the fulfillment of their capabilities and are traceable up to the strategy.
We have a really nice collection of patterns that recognize that the data is the mechanism that ties it together. I would have a look at the open platform and the work they are doing to tie-in the service catalog, which is a collection of activities that human systems or machines need in order to fulfill their roles and capabilities.
The notion of data catalogs, which are the children of these service catalogs, provides the proof of the activities of human systems, machines, and sensors to the fulfillment of their capabilities and then are traceable up to the strategy.
I think we have a nice collection of standards and a global collection of folks who are delivering on that idea today.
Gardner: What would you like to see as a consumer, on the receiving end, if you will, of organizations like The Open Group when it comes to improving your ability to deliver smart city initiatives?
Use-case consumer value
Sunderland: I like the idea of reference architectures attached to use cases because -- for better or worse -- when folks engage around these issues -- even in large entities like New York City -- they are going to be engaging for specific needs.
Reference architectures are really great because they give you an intuitive view of how things fit. But the real meat is the use case, which is applied against the reference architecture. I like the idea of developing workgroups around a handful of reference architectures that address specific use cases. That then allows a catalog of use cases for those who facilitate solutions against those reference architectures. They can look for cases similar to ones that they are attempting to resolve. It’s a good, consumer-friendly way to provide value for the work you are doing.
Gardner: I’m sure there will be a lot more information available along those lines at www.opengroup.org.
When you improve frameworks, interoperability, and standardization of data frameworks, what success factors emerge that help propel the efforts forward? Let’s identify attractive drivers of future smart city initiatives. Let’s start with Dr. Lisdorf. What do you see as a potential use case, application, or service that could be a catalyst to drive even more smart cities activities?
Lisdorf: Right now, smart cities initiatives are out of control. They are usually done on an ad-hoc basis. One important way to get standardization enforced -- or at least considered for new implementations – is to integrate the effort as a necessary step in the established procurement and security governance processes.
Whenever new smart cities initiatives are implemented, you would run them through governance tied to the funding and the security clearance of a solution. That’s the only way we can gain some sort of control.
This approach would also push standardization toward vendors because today they don’t care about standards; they all have their own. If we included in our procurement and our security requirements that they need to comply with certain standards, they would have to build according to those standards. That would increase the overall interoperability of smart cities technologies. I think that is the only way we can begin to gain control.
Gardner: Dr. Harding, what do you see driving further improvement in smart cities undertakings?
Prioritize policy and people
Harding: The focus should be on the policy around data sharing. As I mentioned, I see two layers of a framework: A policy layer and a technical layer. The understanding of the policy layer has to come first because the technical layer supports it.
The development of policy around data sharing -- or specifically on personal data sharing because this is a hot topic. Everyone is concerned with what happens to their personal data. It’s something that cities are particularly concerned with because they hold a lot of data about their citizens.
Gardner: Dr. Saha, same question to you.
Saha: I look at it in two ways. One is for cities to adopt smart city approaches. Identify very-high-demand use cases that pertain to environmental mobility, or the economy, or health -- or whatever the priority is for that city.
Identifying such high-demand use cases is important because the impact is directly seen by the people, which is very important because the benefits of having a smarter city are something that need to be visible to the people using those services, number one.
The other part, that we have not spoken about, is we are assuming that the city already exists, and we are retrofitting it to become a smart city. There are places where countries are building entirely new cities. And these brand-new cities are perfect examples of where these technologies can be tried out. They don’t yet have the complexities of existing cities.
It becomes a very good lab, if you will, a real-life lab. It’s not a controlled lab, it’s a real-life lab where the services can be rolled out as the new city is built and developed. These are the two things I think will improve the adoption of smart city technology across the globe.
Gardner: Don Brancato, any ideas on catalysts to gain standardization and improved smart city approaches?
City smarts and safety first
Brancato: I like Dr. Harding’s idea on focusing on personal data. That’s a good way to take a group of people and build a tactical pattern, and then grow and reuse that.
In terms of the broader city, I’ve seen a number of cities successfully introduce programs that use the notion of a safe city as a subset of other smart city initiatives. This plays out well with the public. There’s a lot of reuse involved. It enables the city to reuse a lot of their capabilities and demonstrate they can deliver value to average citizens.
In order to keep cities involved and energetic, we should not lose track of the fact that people move to cities because of all of the cultural things they can be involved with. That comes from education, safety, and the commoditization of price and value benefits. Being able to deliver safety is critical. And I suggest the idea of traceability of personal data patterns has a connection to a safe city.
Traceability in the Enterprise Architecture world should be a standard artifact for assuring that the programs we have trace to citizen value and to business value. Such traceability and a model link those initiatives and strategies through to the service -- all the way down to the data, so that eventually data can be tied back to the roles.
For example, if I am an individual, data can be assigned to me. If I am in some role within the city, data can be assigned to me. The beauty of that is we automate the role of the human. It is even compounded to the notion that the capabilities are done in the city by humans, systems, machines, and sensors that are getting increasingly smarter. So all of the data can be traceable to these sensors.
Gardner: Don Sunderland, what have you seen that works, and what should we doing more of?
Sunderland: I am still fixated on the idea of creating direct demand. We can’t generate it. It’s there on many levels, but a kind of guerrilla tactic would be to tap into that demand to create location-aware applications, mobile apps, that are freely available to citizens.
The apps can use existing data rather than trying to go out and solve all the data sharing problems for a municipality. Instead, create a value-added app that feeds people location-aware information about where they are -- whether it comes from within the city or without. They can then become habituated to the idea that they can avail themselves of information and services directly, from their pocket, when they need to. You then begin adding layers of additional information as it becomes available. But creating the demand is what’s key.
When 311 was created in New York, it became apparent that it was a brand. The idea of getting all those services by just dialing those three digits was not going to go away. Everybody wanted to add their services to 311. This kind of guerrilla approach to a location-aware app made available to the citizens is a way to drive more demand for even more people.
You may also be interested in:
To head off the threat of food shortages for a global population estimated to top 9 billion by 2050, the world’s agricultural output must double. That mandates innovation to improve monitoring of conditions in the field in order to reduce inputs while maximizing yield and nutritional value. It also means processing data from agricultural land, machines and facilities more efficiently to accelerate research.
These are ideal applications for IoT technologies and edge computing, which is why Hewlett Packard Enterprise is partnering with Purdue University, one of the world’s leading agricultural colleges, to create a new vision for farming and agricultural research in the 21st century. The partnership’s efforts attracted a lot of attention at HPE Discover Las Vegas in June. HPE’s Janice Zdankus, VP for Quality, and Purdue University Executive Sponsor, joined Patrick Smoker, Director and Department Head of Agriculture IT at Purdue, to talk about massive innovation to drive a smarter, more connected, more sustainable agriculture.
Watch the video to learn:
- How edge computing powered by HPE Edgeline and connectivity tech from Aruba, an HPE company, capture terabytes of data from every inch of Purdue’s 1400-plus acre field research station
- How intelligent edge technologies accelerate time-to-discovery for research teams
- How the partners’ innovations will support economic development in Purdue’s home state of Indiana and around the world.
Patrick expanded on these comments in an interview with tech blogger Jake Ludington. How will IoT technologies – including wearables – improve the health and living conditions of livestock? How does the university’s research translate into entrepreneurial opportunities? Watch the video to find out.
Janice also talked with Jake in the interview below. Learn how the partnership with Purdue fits into the broader framework of HPE’s philanthropic efforts, and what comes next for the partners’ digital agriculture initiative.
The Intelligent Edge was one of the main themes at HPE Discover 2018. We announced new edge-to-cloud solutions that enable organizations to run unmodified enterprise-class applications and management software at the edge. Learn more in this post: Unleash the power of the cloud, right at your edge. The latest HPE Edgeline Systems capabilities.
Learn more about HPE Edgeline Converged Edge Systems here.
The next BriefingsDirect developer productivity insights interview explores how a South African insurance innovator has built a modern hyperconverged infrastructure (HCI) IT environment that replicates databases so fast that developers can test and re-test to their hearts’ content.
Here to help us explore the myriad benefits of a data transfer intensive environment is Jacobus Steyn, Operations Manager at King Price in Pretoria, South Africa. The discussion is moderated by Dana Gardner, principal analyst at Interarbor Solutions.
Here are some excerpts:
Gardner: What have been the top trends driving your interest in modernizing your data replication capabilities?
Steyn: One of the challenges we had was the business was really flying blind. We had to create a platform and the ability to get data out of the production environment as quickly as possible to allow the business to make informed decisions -- literally in almost real-time.
Gardner: What were some of the impediments to moving data and creating these new environments for your developers and your operators?
How to solve key challenges
With HPE SimpliVity HCI
Steyn: We literally had to copy databases across the network and onto new environments, and that was very time consuming. It literally took us two to three days to get a new environment up and running for the developers. You would think that this would be easy -- like replication. It proved to be quite a challenge for us because there are vast amounts of data. But the whole HCI approach just eliminated all of those challenges.
Gardner: One of the benefits of going at the infrastructure level for such a solution is not only do you solve one problem-- but you probably solve multiple ones; things like replication and deduplication become integrated into the environment. What were some of the extended benefits you got when you went to a hyperconverged environment?
Time, Storage Savings
Steyn: Deduplication was definitely one of our bigger gains. We have had six to eight development teams, and I literally had an identical copy of our production environment for each of them that they used for testing, user acceptance testing (UAT), and things like that.
At any point in time, we had at least 10 copies of our production environment all over the place. And if you don’t dedupe at that level, you need vast amounts of storage. So that really was a concern for us in terms of storage.
Gardner: Of course, business agility often hinges on your developers’ productivity. When you can tell your developers, “Go ahead, spin up; do what you want,” that can be a great productivity benefit.
Steyn: We literally had daily fights between the IT operations and infrastructure guys and the developers because they were needed resources and we just couldn’t provide them with those resources. And it was not because we didn’t have resources at hand, but it was just the time to spin it up, to get to the guys to configure their environments, and things like that.
It was literally a three- to four-day exercise to get an environment up and running. For those guys who are trying to push the agile development methodology, in a two-week sprint, you can’t afford to lose two or three days.
Gardner: You don’t want to be in a scrum where they are saying, “You have to wait three or four days.” It doesn’t work.
Steyn: No, it doesn’t, definitely not.
Gardner: Tell us about King Price. What is your organization like for those who are not familiar with it?
As your vehicle depreciates, so does your monthly insurance premium. That has been our biggest selling point.
Steyn: King Price initially started off as a short-term insurance company about five years ago in Pretoria. We have a unique, one-of-a-kind business model. The short of it is that as your vehicle’s value depreciates, so does your monthly insurance premium. That has been our biggest selling point.
We see ourselves as disruptive. But there are also a lot of other things disrupting the short-term insurance industry in South Africa -- things like Uber and self-driving cars. These are definitely a threat in the long term for us.
It’s also a very competitive industry in South Africa. Sowe have been rapidly launching new businesses. We launched commercial insurance recently. We launched cyber insurance. Sowe are really adopting new business ventures.
How to solve key challenges
With HPE SimpliVity HCI
Gardner: And, of course, in any competitive business environment, your margins are thin; you have to do things efficiently. Were there any other economic benefits to adopting a hyperconverged environment, other than developer productivity?
Steyn: On the data center itself, the amount of floor space that you need, the footprint, is much less with hyperconverged. It eliminates a lot of requirements in terms of networking, switching, and storage. The ease of deployment in and of itself makes it a lot simpler.
On the business side, we gained the ability to have more data at-hand for the guys in the analytics environment and the ratings environment. They can make much more informed decisions, literally on the fly, if they need to gear-up for a call center, or to take on a new marketing strategy, or something like that.
Gardner: It’s not difficult to rationalize the investment to go to hyperconverged.
Worth the HCI Investment
Steyn: No, it was actually quite easy. I can’t imagine life or IT without the investment that we’ve made. I can’t see how we could have moved forward without it.
Gardner: Give our audience a sense of the scale of your development organization. How many developers do you have? How many teams? What numbers of builds do you have going on at any given time?
Steyn: It’s about 50 developers, or six to eight teams, depending on the scale of the projects they are working on. Each development team is focused on a specific unit within the business. They do two-week sprints, and some of the releases are quite big.
It means getting the product out to the market as quickly as possible, to bring new functionality to the business. We can’t afford to have a piece of product stuck in a development hold for six to eight weeks because, by that time, you are too late.
Gardner: Let’s drill down into the actual hyperconverged infrastructure you have in place. What did you look at? How did you make a decision? What did you end up doing?
Steyn: We had initially invested in Hewlett Packard Enterprise (HPE) SimpliVity 3400 cubes for our development space, and we thought that would pretty much meet our needs. Prior to that, we had invested in traditional blades and storage infrastructure. We were thinking that we would stay with that for the production environment, and the SimpliVity systems would be used for just the development environments.
The gains we saw were just so big ... Now we have the entire environment running on SimpliVity cubes.
But the gains we saw in the development environment were just so big that we very quickly made a decision to get additional cubes and deploy them as the production environment, too. And it just grew from there. Sowe now have the entire environment running on SimpliVity cubes.
We still have some traditional storage that we use for archiving purposes, but other than that, it’s 100 percent HPE SimpliVity.
Gardner: What storage environment do you associate with that to get the best benefits?
Keep Storage Simple
Steyn: We are currently using the HPE 3PAR storage, and it’s working quite well. We have some production environments running there; a lot of archiving uses for that. It’s still very complementary to our environment.
Gardner: A lot of organizations will start with HCI in something like development, move it toward production, but then they also extend it into things like data warehouses, supporting their data infrastructure and analytics infrastructure. Has that been the case at King Price?
Steyn: Yes, definitely. We initially began with the development environment, and we thought that’s going to be it. We very soon adopted HCI into the production environments. And it was at that point where we literally had an entire cube dedicated to the enterprise data warehouse guys. Those are the teams running all of the modeling, pricing structures, and things like that. HCI is proving to be very helpful for them as well, because those guys, they demand extreme data performance, it’s scary.
How to solve key challenges
With HPE SimpliVity HCI
Gardner: I have also seen organizations on a slippery slope, that once they have a certain critical mass of HCI, they begin thinking about an entire software-defined data center (SDDC). They gain the opportunity to entirely mirror data centers for disaster recovery, and for fast backup and recovery security and risk avoidance benefits. Are you moving along that path as well?
Steyn: That’s a project that we launched just a few months ago. We are redesigning our entire infrastructure. We are going to build in the ease of failover, the WAN optimization, and the compression. It just makes a lot more sense to just build a second active data center. So that’s what we are busy doing now, and we are going to deploy the next-generation technology in that data center.
Gardner: Is there any point in time where you are going to be experimenting more with cloud, multi-cloud, and then dealing with a hybrid IT environment where you are going to want to manage all of that? We’ve recently heard news from HPE about OneSphere. Any thoughts about how that might relate to your organization?
Cloud Common Sense
Steyn: Yes, in our engagement with Microsoft, for example, in terms of licensing of products, this is definitely something we have been talking about. Solutions like HPE OneSphere are definitely going to make a lot of sense in our environment.
There are a lot of workloads that we can just pass onto the cloud that we don’t need to have on-premises, at least on a permanent basis. Even the guys from our enterprise data warehouse, there are a lot of jobs that every now and then they can just pass off to the cloud. Something like HPE OneSphere is definitely going to make that a lot easier for us.
You may also be interested in:
- Containers, microservices, and HCI help governments in Norway provide safer public data sharing
- Pay-as-you-go IT models provide cost and operations advantages for Northrop Grumman
- Ericsson and HPE accelerate digital transformation via customizable mobile business infrastructure stacks
- A tale of two hospitals—How healthcare economics in Belgium hastens need for new IT buying schemes
- How VMware, HPE, and Telefonica together bring managed cloud services to a global audience
- Retail gets a makeover thanks to data-driven insights, edge computing, and revamped user experiences
- Inside story on HPC's role in the Bridges Research Project at Pittsburgh Supercomputing Center
- How UBC gained TCO advantage via flash for its EduCloud cloud storage service
- As enterprises face mounting hybrid IT complexity, new management solutions beckon
The next BriefingsDirect agile data center architecture interview explores how an Ericsson and Hewlett Packard Enterprise (HPE) partnership establishes a mobile telecommunications stack that accelerates data services adoption in rapidly advancing economies.
We’ll now learn how this mobile business support infrastructure possesses a low-maintenance common core -- yet remains easily customizable for regional deployments just about anywhere.
Here to help us define the unique challenges of enabling mobile telecommunications operators in countries such as Bangladesh and Uzbekistan, we are joined by Mario Agati, Program Director at Ericsson, based in Amsterdam, and Chris James-Killer, Sales Director for HPE. The interview is conducted by Dana Gardner, Principal Analyst at Interarbor Solutions.
Here are some excerpts:
Gardner: What are the unique challenges that mobile telecommunications operators face when they go to countries like Bangladesh?
Agati: First of all, these are countries with a very low level of revenue per user (RPU). That means for them cost efficiency is a must. All of the solutions that are going to be implemented in those countries should be, as much as possible, focused on cost efficiency, reusability, and industrialization. That’s one of the main reasons for this program. We are addressing those types of needs -- of high-level industrialization and reusability across countries where cost-efficiency is king.
Gardner: In such markets, the technology needs to be as integrated as possible because some skill sets can be hard to come by. What are some of the stack requirements from the infrastructure side to make it less complex?
James-Killer: These can be very challenging countries, and it’s key to do the pre-work as systematically as you can. So, we work very closely with the architects at Ericsson to ensure that we have something that’s repeatable, that’s standardized and delivers a platform that can be rolled out readily in these locations.
Even countries such as Algeria are very difficult to get goods into, and so we have to work with customs, we have to work with goods transfer people; we have to work on local currency issues. It’s a big deal.
Learn More About the
HPE and Ericsson Alliance
Gardner: In a partnership like this between such major organizations as Ericsson and HPE, how do you fit together? Who does what in this partnership?
Agati: At Ericsson, we are the prime integrator responsible for running the overall digital transformation. This is for a global operator that is presently in multiple countries. It shows the complexity of such deals.
We are responsible for delivering a new, fully digital business support system (BSS). This is core for all of the telco services. It includes all of the business management solutions -- from the customer-facing front end, to billing, to charging, and the services provisioning.
In order to cope with this level of complexity, we at Ericsson rely on a number of partners that are helping us where we don’t have our own solutions. And, in this case, HPE is our selected partner for all of the infrastructure components. That’s how the partnership was born.
Gardner: From the HPE side, what are the challenges in bringing a data center environment to far-flung parts of the world? Is this something that you can do on a regional basis, with a single data center architecture, or do you have to be discrete to each market?
Your country, your data center
James-Killer: It is more bespoke than we would like. It’s not as easy as just sending one standard shipping container to each country. Each country has its own dynamic, its own specific users.
The other item worth mentioning is that each country needs its own data center environment. We can’t share them across countries, even if the countries are right next to each other, because there are laws that dictate this separation in the telecommunications world.
So there are unique attributes for each country. We work with Ericsson very closely to make sure that we remove as many itemized things as we can. Obviously, we have the technology platform standardized. And then we work out what’s additionally required in each country. Some countries require more of something and some countries require less. We make sure it’s all done ahead of time. Then it comes down to efficient and timely shipping, and working with local partners for installation.
Gardner: What is the actual architecture in terms of products? Is this heavily hyper-converged infrastructure (HCI)-oriented, and software-defined? What are the key ingredients that allow you to meet your requirements?
James-Killer: The next iterations of this will become a lot more advanced. It will leverage a composable infrastructure approach to standardize resources and ensure they are available to support required workloads. This will reduce overall cost, reduce complexity, and make the infrastructure more adaptable to the end customers’ business needs and how they change over time. Our HPE Synergy solution is a critical component of this infrastructure foundation.
At the moment we have to rely on what’s been standardized as a platform for supporting this BSS portfolio.
This platform has been established for years and years. So it is not necessarily on the latest technology ... but it's a good, standardized, virtualized environment to run this all in a failsafe way.
We have worked with Ericsson for a long time on this. This platform has been established for years and years. So it is not necessarily on the latest technology; the latest is being tested right now. For example, the Ericsson Karlskrona BSS team in Sweden is currently testing HPE Synergy. But, as we speak, the current platform is HPE Gen9 so it’s ProLiant Servers. HPE Aruba is involved; a lot of heavy-duty storage is involved as well.
But it’s a good, standardized, virtualized environment to run this all in a failsafe way. That’s really the most critical thing. Instead of being the most advanced, we just know that it will work. And Ericsson needs to know that it will work because this platform is critical to the end-users and how they operate within each country.
Gardner: These so-called IT frontiers countries -- in such areas as Southeast Asia, Oceania, the Middle East, Eastern Europe, and the Indian subcontinent -- have a high stake in the success of mobile telecommunications. They want their economies to grow. Having a strong mobile communications and data communications infrastructure is essential to that. How do we ensure the agility and speed? How are you working together to make this happen fast?
Architect globally, customize locally
Agati: This comes back to the industrialization aspect. By being able to define a group-wide solution that is replicable in each of these countries, you are automatically providing a de facto solution in countries where it would be very difficult to develop locally. They obtain a complex, state-of-the-art core telco BSS solution. Thanks to this group initiative, we are able to define a strong set of capabilities and functions, an architecture that is common to all of the countries.
That becomes a big accelerator because the solution comes pre-integrated, pre-defined, and is just ready to be customized for whatever remains to be done locally. There are always aspects of the regulations that need to be taken care of locally. But you can start from a predefined asset that is already covering some 80 percent of your needs.
Learn More About the
HPE and Ericsson Alliance
In a relatively short time, in those countries, they obtain a state-of-the-art, brand-new, digital BSS solution that otherwise would have required a local and heavy transformation program -- with all of the complexity and disadvantages of that.
Gardner:And there’s a strong economic incentive to keep the total cost of IT for these BSS deployments at a low percentage of the carriers’ revenue.
Shared risk, shared reward
Agati: Yes. The whole idea of the digital transformation is to address different types of needs from the operator’s perspective. Cost efficiency is probably the biggest driver because it’s the one where the shareholders immediately recognize the value. There are other rationales for digital transformation, such as relating to the flexibility in the offering of new services and of embracing new business models related to improved customer experiences.
On the topic of cost efficiency, we have created with a global operator an innovative revenue-share deal. From our side, we commit to providing them a solution that enables them a certain level of operational cost reduction.
The current industry average cost of IT is 5 to 6 percent of total mobile carrier revenue. Now, thanks to the efficiency that we are creating from the industrialization and re-use across the entire operator’s group, we are committed to bringing the operational cost down to the level of around 2 percent. In exchange, we will receive a certain percentage of the operator’s revenue back.
That is for us, of course, a bold move. I need to say this clearly, because we are betting on our capability of not only providing a simple solution, but on also providing actual shareholder value, because that's the game we are actually playing in now.
It's a real quality of life issue ... These people need to be connected and haven't been connected before.
We are risking our own money on it at the end of the game. So that's what makes the big difference in this deal against any other deal that I have seen in my career -- and in any other deal that I have seen in this industry. There is probably no one that is really taking on such a huge challenge.
Gardner: It's very interesting that we are seeing shared risks, but then also shared rewards. It's a whole different way of being in an ecosystem, being in a partnership, and investing in big-stakes infrastructure projects.
Gardner: There has been recent activity for your solutions in Bangladesh. Can you describe what's been happening there, and why that is illustrative of the value from this approach?
Agati:Bangladesh is one of the countries in the pipeline, but it is not yet one of the most active. We are still working on the first implementation of this new stack. That will be the one that will set the parameters and become the template for all the others to come.
The logic of the transformation program is to identify a good market where we can challenge ourselves and deliver the first complete solution, and then reuse that solution for all of the others. This is what is happening now; we’re in the advanced stages of this pilot project.
Gardner: Yes, thank you. I was more referring to Bangladesh as an example of how unique and different each market can be. In this case, people often don't have personal identification; therefore, one needs to use a fingerprint biometric approach in the street to sell a SIM to get them up and running, for example. Any insight on that, Chris?
Learn More About the
HPE and Ericsson Alliance
James-Killer: It speaks to the importance of the work that Ericsson is doing in these countries. We have seen in Africa and in parts of the Middle East how important telecommunications is to an individual. It's a real quality of life issue. We take it for granted in Sweden; we certainly take advantage of it in my home country of Australia. But in some of these countries you are actually making a genuine difference.
These people need to be connected and haven’t been connected before. And you can see what has happened politically when the people have been exposed to this kind of technology. So it's admirable, I believe, what Ericsson is doing, particularly commercially, and the way that they are doing it.
It also speaks to Ericsson's success and the continued excitement around LTE and 4G in these markets; not actually 5G yet. When you visit Ericsson's website or go to Ericsson’s shows, there's a lot of talk about autonomous vehicles and working with Volvo and working with Scania, and the potential of 5G for smart cities initiatives. But some of the best work that Ericsson does is in building out the 4G networks in some of these frontier countries.
Agati: If I can add one thing. You mentioned how specific requirements are coming from such countries as Bangladesh, where we have the specific issue related to identity management. This is one of the big challenges we are now facing, of gaining the proper balance between coping with different local needs, such as different regulations, different habits, different cultures -- but at the same time also industrializing the means, making them repeatable and making that as simple as possible and as consistent as possible across all of these countries.
There is a continuous battle between the attempts to simplify and the reality check on what does not always allow simplification and industrialization. That is the daily battle that we are waging: What do you need and what don’t you need. Asking, “What is the business value behind a specific capability? What is the reasoning behind why you really need this instead of that?”
We at Ericsson want to be the champion of simplicity and this project is the cornerstone of going in that direction.
At the end of the game, this is the bet that we are making together with our customers -- that there is a path to where you can actually find the right way to simplification. Ericsson has recently been launching our new brand and it is about this quest for making it easier. That's exactly our challenge. We want to be the champion of simplicity and this project is the cornerstone of going in that direction.
Gardner: And only a global integrator with many years of experience in many markets can attain that proper combination of simplicity and customization.
You may also be interested in:
- How VMware, HPE, and Telefonica together bring managed cloud services to a global audience
- IoT capabilities open new doors for Miami telecoms platform provider Identidad IoT
- How Nokia refactors the video delivery business with new time-managed IT financing models
- Retail gets a makeover thanks to data-driven insights, edge computing, and revamped user experiences
- As enterprises face mounting hybrid IT complexity, new management solutions beckon
- How a large Missouri medical center developed an agile healthcare infrastructure security strategy
- Get ready for the Post-Cloud World
- Philips teams with HPE on ecosystem approach to improve healthcare informatics-driven outcome
- Inside story: How Ormuco abstracts the concepts of private and public cloud across the globe
- Inside story on developing the ultimate SDN-enabled hybrid cloud object storage environment