Thought for the day: The invisible computer

'Smart' computers will not only save us time, they will save us money and offer businesses greater flexibility too, says Alan...

'Smart' computers will not only save us time, they will save us money  and offer businesses greater flexibility too, says Alan Ganek.

Consider this fact: between a third to a half of a company's total IT budget is spent preventing or recovering from crashes, according to a recent study by the University of California.  And no wonder.  A system failure at a financial firm or online retail store can run up millions of dollars in lost business.

And, if you look at computer outages, about 40% are caused by operator error. That's not because operators aren't well trained or don't have the right capabilities. It's because the complexities of technology are difficult to figure out, and IT managers are under pressure to make decisions in seconds.  

Businesses can't just roll in processors and storage fast enough to avoid meltdowns when usage spikes, or fend off viruses and hacker attacks, or manage the different operating systems that must plug in and access information: laptops, palmtops, desktops, whatever.  People are good, but they're not that good.  

Since no one is close to writing defect-free software or hardware, what we need are computers capable of running themselves and with far greater levels of intelligence built right into the technology itself.  

And by "intelligence", I'm not talking about computers that can write the next Ninth Symphony.  I'm talking about the same kind of intelligence we take for granted in our own bodies.  We walk up three flights of stairs and our heart rate increases.  It's hot, so we perspire.  It's cold, so we shiver. We don't tell ourselves to do these things. They just happen.  We need something similar for business.

If systems and networks could begin to adopt these attributes, managers would be able to set business goals and computers would, automatically, set the IT actions needed to deliver them. For example, in a financial trading environment, a manager might decide that trades have to be completed in less than a second to realise service and profitability goals. It would be up to software tools to configure the computer systems to meet those metrics.

The implications for this "autonomic" business approach are immediately evident: A network of organised, "smart" computing components that give us what we need, when we need it, without a conscious mental or even physical effort.

Here's an example. A computer freezes up intermittently, no customer transactions are being processed for several seconds, losing thousands of dollars in business and customer confidence and loyalty.  Today, IT support staff might not even find out about the problem for more than a day. When they do, it takes a couple days to figure out what is happening.  

With an autonomic system in place - with real-time monitoring and auto-tuning analysis - the freeze-up is detected the first time it happens and matched against historical problem data. The settings are reset automatically, averting any real loss of revenue and customer loyalty. A report that the administrator reads the next day shows the action that was taken.  

The good news is work is already being done to make this happen. There are research projects at labs and universities that include self-evolving systems that can monitor themselves and adjust to changes and "cellular" chips capable of recovering from failure.  

The goal is to increase the amount of automation businesses have.  Because the more that you reduce human error, the more efficient your business will be - whether you're a financial institution, a shipping company or an online retailer. The beauty of it is that all of the complexity gets hidden from the user.

The logic is compelling: relief from the headaches of technology ownership and maintenance; an improved balance sheet; and much greater flexibility in meeting the demands of running a business.  

But in the end, perhaps the greatest benefit would be the freedom it would unlock.  Sure, it will create enormous efficiencies.  But the game-changing impact will be freeing up all companies - whether just starting out or well established - to experiment, be more creative, rethink assumptions and stretch the limits of progress as far as the mind can conjure.

What do you think?

Do you think today's computer systems are too complex to manage? Tell us in an e-mail>> reserves the right to edit and publish answers on the website. Please state if your answer is not for publication.

Alan Ganek is vice president of autonomic computing at IBM Software Group.

Read more on Managing IT and business issues