Sensors: the unsung heroes of digital transformation
01 October 2019
As the digital economy has grown over the last few decades, data has become a key commodity – to be captured, funnelled through wireless networks & processed by high-performance supercomputers into valuable intelligence that can be shared & displayed across numerous interconnected devices.
This article was originally featured in the October 2019 issue of EPDT magazine [read the digital issue]. Sign up to receive your own copy each month.
This is made possible by an explosion of innovation in software and hardware, with the emergence of next generation cloud technologies and the Internet of Things (IoT). But often overlooked, says Matthew Ashton, Divisional Manager – Components at optoelectronic specialists, Pacer, are the humble sensors sourcing this data. For example, optoelectronic sensors – acting as the digital drills extracting this valuable commodity.
Interfacing with the latest software, these advanced sensors make new products possible, as well as adding new intelligence and functionality to existing technology – such as the new wave of gesture recognition that promises to make electronic devices everywhere easier to use.
“The sensors have been around for ages” says Pacer’s James Woodhead, “but finally the software has caught up, and the processing power is now sufficient, making it cheap enough to roll out hand gesture recognition technology into industry standard products, from cars, to TVs and vending machines.”
But while innovations in semiconductors and software have made technologies like this cheap enough and fast enough to be widely deployed, they are still ultimately dependent on the accurate generation of data from simple sensors.
The risk of unreliable data
With mass produced sensors readily available at low cost, companies today have more opportunities than ever to improve products by adding new functionality to score a quick win over competitors.
But underestimating the role of sensors – whether by skimping on price or mismatching specifications with usage – reduces their ability to gather data accurately and consistently, undermining the resulting product.
The latest wave of smartwatches and fitness wearables provide a pertinent example.
These wearable technologies offer an age-old product with an added twist. By combining optoelectronic sensors with small specialist custom built chips which require little space and no additional circuitry, manufacturers have been able to create watches and bracelets that claim to be able to measure the number of steps you take, the quality of your sleep and your heart rate – data which is gold dust to health fanatics and athletes, opening up a whole new market of potential consumers.
But, while the sensors themselves may be straightforward – with a typical optical heartbeat sensor detecting the pulse by simply shining a light through the skin to see blood flow – integrating these sensors into the product can be more complex. As Ashton says, it’s not just a matter of “taking the [sensor] output and plugging it into the system – even for newer sensors with a digital output, rather than analogue.”
“Loads of companies are producing pulse oximetry sensors designed to go into these heart rate monitors, but the accuracy of some of these is so atrocious that they may provide a heart rate reading at 150 bpm when the actual rate is 98 bpm. This needs to be acknowledged and duly considered.”
Integrated without an understanding of their design and function, optical heartbeat sensors can suffer from signal loss or corruption, masking the signal from the pulse and producing what is ultimately unreliable data.
One top fitness wearable manufacturer is facing a lawsuit after trials found the product to be “wildly inaccurate” when measuring activity at high intensity levels, potentially leading athletes to damage their fitness by training at a lower intensity than needed, or worse, inducing a heart attack as they push themselves beyond their limits.
The humble sensor
Such situations might be problematic for amateur athletes, but in healthcare – where decisions could mean life or death – getting clear and accurate data from sensors is even more critical.
As in other industries, the convergence of several technological developments has now created new possibilities for the processing of data in healthcare and created higher levels of efficiency than ever before.
Data generated by optoelectronic sensors in a patient’s home might now be pushed up to the cloud, where it can be processed and interpreted by powerful software, and trigger the necessary actions and alerts on doctors’ handheld devices, which themselves rely on the longevity of lithium-ion batteries that have only been developed in the last decade.
Given the growing need for dementia care and the continuous drive to get people out of hospital beds and recovering in the comfort of their own home, this successful capturing and processing of data is increasingly important, and hinges on the motion and light sensors that detect something as simple as a person falling out of bed, leaving a door open or turning a light on.
Having this information about a person’s daily movement can make it feasible for a patient to leave the hospital, safe in the knowledge they will be closely monitored.
The systems that make this possible however, says Ashton, “are only as good as the data that is going into them. If, for example, the patient hasn’t moved for two hours, the sensor must reliably send that data to the web so that it can set off a chain reaction and raise the alert with a doctor or a relative’s smartphone somewhere else. If the sensor is not working correctly or is being confused by something else, then these systems are pointless.”
To avoid the fate of the fitness bracelet, sensors in these systems and others need to be implemented with a good understanding of their specification and purpose – what type of sensor is best to use, where they fit and how theyshould interface with other components.
Designed appropriately, these sensing elements can deliver data to a newly powerful digital infrastructure, unleashing new possibilities across a range of industries, from medical to military and consumer goods – but they must be good quality, accurate and reliable.
However, Ashton suggests, as ever, “the quality of information that you get out depends on the quality of information that is put in”, and unless expertise is applied to the integration of sensors, they’re unlikely to fulfil their role successfully in the digital world. A role which, though often underestimated, is foundational: “without that cornerstone, none of this works.”
Contact Details and Archive...