Teaching systems to make their own decisions
Professor David Sanders' research is opening up a new world of possibilities for children with disabilities
Combining systems engineering with artificial intelligence (AI), he automates decision-making so systems can ‘think’ for themselves. Along the way, he’s providing solutions across a wide range of industries.
David’s pioneering research is opening up a new world of possibilities for children with disabilities.
Working in collaboration with Chailey Heritage in Sussex, David is working on intelligent powered wheelchairs. These will enable children with high levels of disability to drive their own wheelchairs. Some for the very first time.
David’s systems help the children take their wheelchairs where they want to, without colliding. But this is where it gets complicated. David explains:
‘Ever been driven mad by a computer trying to “help” you? Well, something similar happens if you build a system for a wheelchair that avoids collisions.
‘It succeeds in not crashing. But can prevent the user from switching the light on, or putting something in the bin. Because the powered wheelchair won’t let them go to the wall or bin - to avoid a collision.’
This is where the combination of AI and decision-making makes the difference. David’s research looks at how the AI can interpret what the person wants to do, or is trying to do. And then it helps them to do it:
‘But this isn’t easy as humans can be so unpredictable!’
Some of these children have never been able to move independently. For them, David’s research is genuinely life-changing.
‘The social impact of powered wheelchairs on the disabled kids is pretty powerful. It’s very empowering to be able to initiate your own movement rather than being moved around in your wheelchair by someone else.’
David’s work on systems engineering has other applications too. Working with InTandem Systems Ltd at Chichester Festival Theatre, he’s looking at building energy management systems. The goal is to see how much energy they consume in heating and lighting, and find ways to reduce this.
He’s also using intelligent monitoring to improve efficiencies with compressed air systems in industrial and manufacturing settings. If there is a leak in a compressed air system, often all that happens is the pressure is turned up. This wastes huge amounts of energy. So David is building systems to detect these leaks before they become an issue.
And this has benefits all round - including improved efficiencies and cost savings for businesses, as well as lower CO2 emissions through reduced energy usage.
The social impact of powered wheelchairs on the disabled kids is pretty powerful. It’s very empowering to be able to initiate your own movement rather than being moved around in your wheelchair by someone else.
Making sense of it all
Intelligent sensors are proving useful in all kinds of ways - from ensuring a system is running smoothly, through to preventing industrial catastrophes.
David is working on intelligent sensors with Gems Sensors & Controls Ltd in Basingstoke. These can be used to monitor factors including fluid levels and flow, temperature and pressure. And more sensors mean more information. For example, multiple sensors can be used to monitor the amount of fluid going into a system, the amount going out and the overall level. And if these don’t add up, the sensors can highlight that something is wrong.
These intelligent sensors might be able to tell whether the system as a whole is functioning as it should be. And the sensors monitor themselves too. They know whether they are they operating correctly or if they need replacing soon.
Multiple intelligent sensors are particularly useful in avoiding catastrophe - vital in the petroleum and chemical industries. Using multiple inputs to monitor systems provides information on how the whole system is working. And looking at the whole story, rather than individual elements, helps avoid leaks and explosions.
David’s research, translated into innovation, shows that using AI to let systems make their own decisions, using informational data, makes for better decision making.