Accenture's king of blue-sky thinking

Research director Martin Illsley discusses how utility companies could use Flickr to cut costs and how Evel Knievel relates to strategic technology planning

Some companies profess to put out-of-the-box thinking at the heart of their operations. Apple and Google, for example, claim to thrive on leaving no stone unturned when it comes to innovation. For instance, the search giant tries to ensure its developers have 20 percent of their time available for pursuing personal projects, even though they might not come to anything.

But being disruptive and innovative is easier in some industries than others. Financial services is one place where innovation equals risk, which is usually best avoided. However, even those companies in that sector which have embraced a so-called "culture of innovation" still face the dilemma of choosing which areas of technology to focus research efforts and resources on.

While Accenture is best known for its IT consultancy work, the firm also invests in research and development. Martin Illsley, director of research at Accenture Technology Labs, based in Sophia Antipolis, France, spoke to about innovation and some of the big developments in technology that should be on every business's radar.

Q: What are the main challenges facing businesses at the moment in terms of practical application of technology?
A: There are choices dragging people apart about how to balance the physical and the virtual. Bank branches are back but, on the other hand, the virtual is going to the extreme — look at Second Life. Assuming you can do everything virtually, what do you choose to do virtually and what do you choose to do physically? With technology, we have the choice to be antonymous. With inventories, you could put an RFID tag on every chair in the building and, at the press of a button, you could do the inventory. You can drive a car which has a bleeper on the front that lets you go through toll gates at the moment on the continent. On the other end of the scale, mass people participation is important — wikis and blogs, for example. You can have everyone involved, from the policies you put out, to product design, customer feedback and service.

Could you expand on that?
Amara's Law, which also goes by other names, is that people overestimate the short-term effects of an action or technology, while underestimating the long-term effects. Evel Knievel is a case in point of overestimating the short term: he thought he could jump 27 buses and couldn't. We do the same with respect to new technologies; we think we've done it all and that it's going to happen tomorrow. We also underestimate the long term, like Evel Knievel. I'll bet he'd never have guessed that by now he would have given up riding motorcycles and become very religious. He's now riding pillion with God.

It's the same with technology; we underestimate the long-term effects. Take the iPod, for example. We still haven't fully understood its effect on the production of music — it has huge longer-term implications. RFID is not just about tags; it has huge implications for how businesses operate, and how society accepts the potential of being tracked. It's the same with the iPhone. Sure, at the moment, it's being hyped up to fever pitch about what it can do, but the longer-term impact will probably be quite profound. Whether it will lead to a completely convergent device — who knows? It may have an impact on whether people want specialist devices like cameras.

What do you think are going to be the main trends when it comes to development of technology over the next five years?
Firstly, we predict more use of virtualised infrastructures. If you have infrastructures like data centres, you have a choice. You can go to companies like Google and Amazon and buy the capabilities — buy the data centre by how much you use. It's the same with web capabilities. With hosting organisations like 1&1, you get part of a machine somewhere, and they host on a scale that's very economical. Organisations have to think: do we do it ourselves, or buy it? Google is doing virtualised infrastructure on an unprecedented scale, with a proprietary approach that spreads the load over thousands of machines. Many organisations won't have the scale to support it, and virtualised infrastructure will move up the value chain. Infrastructure is moving from cables to machines, systems and applications; organisations can add more on top.

What is your view on the current trend of governments and businesses demanding more from software vendors around interoperability?
We predict seamless interoperability and process-centric IT. The Business Process Execution Language [BPEL] formally specifies what processes are. A flow of processes translates into service-oriented architecture functions. The effect is that we will start to..., contrast, and manage at process level, like a time and motion study. We will manage at this level, rather than manage at the IT-function level.

We also predict closed loop analytics. Where we're at today, we can do predictive analytics. We can predict what will happen with equipment. However, the key aspect is integration. It's no use knowing if a machine is going to break down in three days' time if you haven't organised getting parts or a maintenance crew. More subtly, if you have limited resources, what is the best course of action to take if you have a situation where this machine will break down in two days, this in one, this in five? It's like when aircraft break down: do you put the passengers in a hotel; do you try to repair the aircraft; what are the consequences of each course? It can get very complicated.

One of Accenture's areas of research is communications. What are your thoughts on convergence?
We predict fluid collaboration platforms. In some organisations you have videoconferencing rooms that lie empty, then some conference calls with 20 or 30 people, and nothing gets done. There's nothing wrong with the tools, there's just no formalisation of the process of inviting the right people integrated into systems. If the object of a meeting is to make certain decisions, tools should make sure the right people are there.

Mobility for enterprise applications will also become important. There's a trend for consumer usage, but there's a lot more that can be done on the enterprise side. As we get 3G video capabilities, there are large implications for business — a field force could send pictures and videos to each other. But people haven't thought through the implications of how to deal with video in a call-centre way.

What can businesses gain from Web 2.0? Should businesses be concerned about the proliferation of new applications that have to be assimilated into their systems?
Web 2.0 has two parts. With end-user computing, it's scary how many mashups are appearing — it reminds me of PowerBuilder, which allowed business users to write and build reports. The IT department had to manage these applications, and maybe didn't know where or what they were. If business users are now creating non-standard applications, the poor old IT department has to manage this. Mashups are a powerful capability, but businesses have to be careful. It's the same with socialising.

What are your views on customer discussion in online forums?
There are lots of benefits of having online forums where customers can hammer out solutions to issues, but this can have a negative side. If there is an issue, it becomes an open, visible thing. If it's not true, it's even worse. It's a matter of getting the balance right.

User content can be used — YouTube and Flickr get tons of content from people. If you're a utility company, you could set up a mass-participation site where people send in pics of damage to infrastructure, cutting down on inspection costs. Okay, 70 percent of the pictures may be rubbish, but 30 percent will be useful. It can be a powerful tool with useful cost benefits. However, it's also publicising the damage.

Organisations need to classify content. They tend to be conservative; they don't go through their data to tag corporate information. They need to define which information can be talked about with which groups. They have to do that before they are too abundant with throwing out information.

Are there other Web 2.0 technologies that could be disruptive?
We predict a greater use of widgets — internet applications with a simple interface that are not part of the browser. We've invented a widget that can be back-ended to an enterprise resource planning system, [and] that can then drag out information and push it onto a mobile device. There's a lot of scope in that — giving access to business systems through this type of technology. As the mobile widgets are sandboxed on the platform, security is not an issue.

What other blue-sky areas of research is Accenture involved in?
The industrialisation of software development. We're interested in code that, with a slight modification, can change itself. At the moment we write code and have to throw it away and build from the ground up to put in another bit. Ultimately we'd like to have code development done only once, then it will handle any changes that need to be made itself. When you build a house, you don't want to knock it down and start again every time you want to make a change, and it should be the same with code development.

So you're talking about code with artificial intelligence capabilities?
Not exactly, but code could track whether it is used or not and, if it wasn't used over time, it would self delete, using an intelligent approach to self correct. This would require a more formal description of what software is supposed to do. You could also monitor services — which are being used and which aren't.