It is something all of us have experienced. You pick up the phone, and suddenly nothing. The screen has switched off as you’ve run out of batteries. It’s as frustrating as it is annoying. However, for remote monitoring systems and embedded sensors this annoyance becomes a costly maintenance procedure, as batteries need to be changed and charged, with routine scheduling.
For this reason, Innovate UK set up the Energy Harvesting Special Interest Group (EHSIG), which aims to develop techniques to harness energy from the ambient environment. The term ‘energy scavenging’ is often interchangeable used and describes various methods of capturing thermal and kinetic energy from the surrounding environment to power small autonomous sensors.
Dr Alex Efimov, a senior academic researcher for Innovate UK, has been involved with the EH SIG since its inception. He said: “The boundary we set for something to be considered ‘energy harvesting’ is anything under one watt of power, so that means solar panels aren’t considered the same thing.
“The idea behind energy harvesting is that they will one day replace batteries all together in low-power applications. No one really likes batteries as they are consumable. You have replace them, and that all has a cost. The cost of replacing batteries is often way more substantial than the actual cost of the batteries themselves.”
Yet, in practice, batteries are commonplace as engineers know what it is they are getting in terms of controlled electrical power output. Energy harvesting, however, is the polar opposite. Power can vary greatly, is reliant on external inputs and the environment, and as yet, have not found a great deal of practical application.
“It is difficult to find a common denominator and come up with a single energy harvester that works for every application,” explained Efimov. “There are different technologies and perhaps you might combine them to answer a specific need. This is where designers and system integrators are very important, as they know how to put it all together so it all works in the end. But, building an energy harvesting business out of that is difficult, as there are so many energy harvesting technologies that fit many applications.”
Bucking this trend is Southampton based Perpetuum. The company commercialised a vibration harvesting system in 2004 to power vibration and temperature sensors placed on train boogies to monitor a number of parameters from, bearing wear to excess heat in the brakes. It’s allowed a move to real-time condition monitoring and proactive maintenance.
The wireless sensor nodes completely do away with batteries, and instead use a patented electromagnetic-energy harvesting mechanism to convert mechanical energy produced by vibrations on the train in to electrical energy. This in turn is enough to power the wireless sensor nodes that can monitor the ball bearings in the wheels, and transmit this real-time data to the cab where it can be stored and later put through comprehensive analytics.
The configuration has proven to be a reliable fit and forget solution. Indeed, the operational life of the energy harvesting part of the device is over 100 years, and the sensing nodes themselves 20 years. The data that is captured can be communicated over a long range through GPRS transmissions.
Justin Southcombe, commercial director at Perpetuum, said: “Our system is giving 11,000 snapshots a day on the Southeastern network so we can tell where the outliers are... This means we can track the rate of change on, for example, a set of points.
“We shouldn’t be running a 21st century railway where we measure the state of the track through the driver’s backside. The driver could feel a bump, which might just be some stray ballast on the railhead – but a track team has to go out and take a look. We can see if that area has been trending: did all the passing wheels on that day feel a bump at that spot? If they did, it could be something that requires attention to avoid having to impose a temporary speed restriction.”
Ideal application
While probably the forefront example of a commercialised energy harvesting sensor system, it is by no means the only one to have been developed. A team of engineers from Cambridge University are in the process of deploying a self-powered wireless sensor network to monitor the Forth Road Bridge in Scotland.
The team from the Cambridge Centre for Smart Infrastructure and Construction (CSIC) has designed vibration energy harvesters that again convert the ambient vibrations of the passing traffic into electricity, to allow numerous sensors placed on the bridge to monitor key parameters on the long-span suspension bridge.
The 2.5km Forth Road Bridge, which connects Edinburgh and Fife, now carries far more traffic than it was originally designed for; about 25 million vehicles a year and nearly ten times the number it carried when it opened in 1964. As a result the increased strain needs to be monitored on a continual basis.
Dr Yu Jia and Dr Ashwin Seshia developed the vibration energy harvesters based on a phenomenon known as parametric resonance, which amplifies the vibrations. It means the devices have the potential to harvest significantly more energy from vibrations than previous designs, and vibration data collected from the Forth Road Bridge during a field investigation is now being used to optimise the harvester further.
Professor Middleton said: “As energy harvesting improves and the amount of energy available to power sensors increases, new radio technologies will emerge with even lower power requirements. We may be approaching the point at which a vibration powered wireless sensor network, with no need to change batteries, becomes a reality.”
Future trends?
The Internet of Things has been a popular topic for engineers as sensors, data capture and then transmission on the mass-scale begins to be implemented. However, the power requirement for many devices and applications is still holding many potential ideas back. Changing batteries, or even charging them, is simply not viable if intelligence in everyday items on the scale that is proposed by many is to become a reality. The Internet of Things needs autonomous sensors that produce enough power to capture and transmit data to truly be rolled out on the scale that many proponents are claiming.
“The human bodies temperature is 36°C, let’s say ambient temperature is on average 20°C,” explained Efimov. “That is a pretty small temperature gradient so you can only reasonably produce micro-amps and fairly small voltages, and that is not a huge amount of power.”
The same goes for vibration harvesting and solar energy harvesting – despite the potential, it is difficult to produce significant quantities of power. So for some time to come, it is unlikely that a mobile phones or even a wearable health monitoring electronic devices will be able to be self sufficient when it comes to its power requirements.
“The real value proposition is replacing batteries in small sensing networks and minimising cost by not having to service them,” said Efimov. “But, I’m sceptical about mobile phones.”
