PUBLIC sector agencies that roll out large-scale infocomm technology (ICT) projects should take a diverse approach or run the risk of being locked into obsolete technology infrastructures over time, said Kishore Swaminathan, chief scientist of global consulting firm Accenture.

Speaking to BizIT yesterday at Accenture's annual seminar for chief information officers (CIOs) and IT managers, Dr Swaminathan said: 'The hard problems with society-wide infrastructure is that while you have to plan ahead, you also have the risk of creating a legacy problem with fast-changing technologies.'

His answer to technology obsolescence: Don't get wedded to any one technology.

In other words, he said: 'Diversify, even though it is very attractive to say that you have the whole place that's based on WiMax, because you may very well create a legacy problem since technology is (always) changing.'

Dr Swaminathan's advice will be particularly relevant for Singapore. Public sector ICT procurement activities have been bullish in the past few years and continue to grow.

One notable upcoming nationwide infrastructural ICT project in Singapore is the Next Generation National Broadband Network, which is slated for nationwide deployment in 2015 and will offer broadband speeds of 1Gbps and beyond.

Another notable infrastructure project was launched in March by the Maritime and Port Authority of Singapore (MPA) and IDA, called Wiseport. Based on the wireless connectivity technology WiMax (Worldwide Interoperability for Microwave Access), it provides ships near Singapore's port with wireless broadband access.

In the case of Wiseport, Dr Swaminathan said that betting on WiMax poses little risk because the scale of this project is relatively small. Where legacy becomes an issue is when the equipment owned by 'thousands and thousands of end-users' have to be changed, he said.

'If you have 8 million citizens and you've bet on the wrong technology, you could have the legacy problem for 10 to 15 years.'

Dr Swaminathan added that public sector ICT planners should look to forecast 'not for today's capacity, but for more capacity than any use you can think of today'.

'When it comes to planning for capacity, humanity generally (under-provision) - if we want to build a bridge across a river, we count the number of people who (have been) swimming across that river and then build a bridge to carry that capacity.

'But after you've built the bridge, there will be 10 orders of magnitude more than the current traffic that will be crossing the river, because enablement increases traffic and use dramatically,' he said.

As to which technologies ICT planners and CIOs should bet on in future, Dr Swaminathan outlined eight key trends in his presentation at the seminar. The first three trends are utility-based - so-called 'cloud' - computing, the parallel emergence of heavy-weight service oriented architectures and lightweight 'widgets' application development approaches, and pervasive enterprise intelligence.

The other five trends are always-on user connectivity, social computing, user-generated content, cheaper and better software development processes and green computing.

This article was first published in The Business Times on May 22, 2008

Is this article useful to you?