Cognitive computing has the promise to become a major asset in helping councils to work smarter, says Mark Say.

Local government isn’t going to get any easier. If the current austerity programme ever comes to an end, an ageing population and a lot of people on low incomes are going to maintain the pressure to do more with less. Councils have got to get smarter in the day-to-day and strategic elements of delivering services.

It’s a daunting prospect, but some eyes are looking at the emergence of cognitive computing for solutions. It’s a few years from realising its promise, but it should be as big a step in the evolution of government as it will be for technology.

Cognitive systems are programmed to use algorithms to continually learn from the data they receive, making them more flexible and intuitive than traditional computers. They come closer to thinking like people, while being able to deal with massive volumes of information much faster and more consistently than the human brain.

In routine processes they can respond to the data, anticipate problems, refine the way they work and save themselves from downtime. For more complex tasks they can accelerate learning to support people making critical decisions, at an operational level and in long term planning.

Both of these hold promise for local government, especially as it looks at tapping into the “internet of things’, the myriad of sensors and mobile devices that generate torrents of data. Cognitive systems can find patterns in that data and provide support for immediate actions and strategic thinking.

IBM, one of the flag wavers for cognitive computing, has highlighted local government as one of the areas in which it can make a big difference, pushing the concept of the ‘smarter city’. It gives examples of how it can make operations smarter. For example:

• sensors can alert waste disposal teams of where there are bins or skips that need emptying;
• cognitive systems can review thousands of CCTV images to spot anything suspicious and alert the police;
• Feeds on weather, traffic and medical services can provide early alerts of a civic emergency.

Looking to the long term, the technology can support planning to deal with some of the big pressures on society, often bringing local authorities together with other agencies. An example is in looking at what contributes to a health problem, such as child obesity, by gathering data on all the factors that could have an effect .

These might include the proximity of fast food outlets to schools, whether there are safe playgrounds and the transport options for taking part in organised sport. There’s little discipline in all this data, and a cognitive system can refine and correlate it to run simulations on more coordinated services to support public health.

It’s a view that has supporters in UK authorities. Jos Creese, chief information officer of Hampshire County Council, says the potential is particularly strong for care services, especially when the broad data is aligned with the circumstances surrounding an individual.

“It’s in making  a connection between the circumstances of the individual – everything from geography, medical condition, the services they’re currently using – in order to make intelligent connections for them to give better service outcomes,” he says.

“The crossover might be between health and social care where they could move to caring agencies or other services.

“At the moment the internet is brilliant, but you have to do the leg work, and trying to make sense of the vast amount of resources is where public services could benefit enormously from cognitive systems.

“We expect people to make connections, or we make the obvious ones when you leave hospital to social care, or from social care to an association for self-help.

“Now we need to make sure those connections are joined up with a judgement system that says a person would really benefit from a specific package of information, services and support.”

On a broader front, there is a lot of potential in helping to get more from the transport, communications and energy infrastructures of communities, and in contingency planning for responding to emergencies.

“The sustainable use of the resources that exist in a city is a massive societal challenge in which cognitive systems can help,” says Rashik Parmar, president of the IBM Academy of Technology.

“These systems can help to optimise the use of the infrastructure and support the move towards smart cities.”

Some questions remain around the limits of using cognitive systems to support critical decisions. There are areas such as traffic and emergency planning, in which the cost of getting things wrong will probably be too high for authorities to take the risk of reducing human experts to a minimal role.

“It depends on the application,” Creese says. “For some, where making a mistake doesn’t have monumental consequences, and if on average you’re getting a better decision-making process with better outcomes, take the risk.

“If, on the other hand, the level of impact is such it could result in life and death you’ve got to be pretty confident before taking the risk.”

So the odds are on cognitive systems supporting, not replacing people in making critical decisions. Authorities will have to decide on what is appropriate, although it is possible that, as the use of the technology increases, central government decides the risks are great enough to make a human decision mandatory for some functions.

It will be a few years before these questions really begin to press; cognitive computing is on the horizon, but it is likely to be up to 10 years before it has a major presence.
But local government will have to develop services that are more efficient and produce better outcomes, and a technology that makes it smarter will be a major asset.

Mark Say is author of Riding the Next Wave – Cognitive Computing, a white paper produced for BCS, The Chartered Institute for IT.