Bench Talk

One of the great things about being human is man’s unfailing curiosity and desire to solve problems. Even though sometimes that ability can take us a bit too far so that we end up with things that really don’t get us anywhere. While humans may be good at solving problems, they’re also pretty good at reinventing the wheel, or even at solving one problem only to find they’ve created another. You only have to look back through history to find it littered with examples.

Some say that’s just what’s happening with the Internet of Things (IoT), to which anywhere between 26 billion and 200 billion devices will be connected by 2020, depending on whose figures you believe. It will give us myriad devices connected together, busily exchanging data like there’s no tomorrow.

The aims for the Internet of Things are altruistic enough. Within the health and wellbeing sector, there’s a lot of talk of using IoT to improve outcomes for patients as much in terms of early diagnosis or prevention of conditions as of helping those with existing conditions to manage them better.

Deloitte Consulting says that the key to this lies in Patient-Gathered Data (PGD). They’re not talking about data reported by patients themselves, which is notoriously unreliable – who hasn’t rounded up figures to make their performance look better or salve their conscience at not having done quite as much as they should have done to achieve a goal, or even understated something that could possibly be viewed as risky behaviour? Rather, this is data originating from patients’ medical devices or wearables and transmitted using mobile devices or patients’ own PCs.

Glucose monitors or fall detection devices are typical of the kind of device being connected up to systems to help patients manage their chronic conditions or increasing age from within their own homes. IoT holds out the promise of delivering huge cost savings in terms of managing chronic disease and improving health outcomes for patients. It should also be invaluable for helping people adopt and maintain healthier lifestyles – key to having a positive effect on individuals’ health and so improving their chances of living longer and more healthily into the future.

The technology looks even more attractive when it is employed in areas of the world with heavily rural and widely disseminated populations with historically low levels of access to medical services. Over the past 20 years or so, greatly improved telecommunication services have enabled increasing numbers of patients to benefit from the expertise of health care professionals situated remotely. So-called telemedicine is used to identify, diagnose and treat conditions in patients. Photos taken by a smartphone can even be used to diagnose illnesses.

Increasingly, here is a golden opportunity for a new breed of IoT device, the senstroller, which combines sensor and microcontroller to gather data and / or control and disseminate it. And here we start getting to one of the key points about IoT devices: the need for them to operate over a wide range of protocols, be able to use a plethora of different gateways and operate alongside senstrollers that are providing different services.

So that’s one design headache – or, rather, opportunity. Back in 2015, the ITU Radiocommunication Assembly established the roadmap for what it’s calling IMT (International Mobile Telecommunication)-2020, better understood by most of us as the development path for 5G technology. This shows us the shape in which future applications (including healthcare) will sit, work and communicate alongside a raft of others. Some will have heavy data requirements and a need for instantaneous transmission; others, like many IoT applications, will carry a much lighter requirement. The network of 2020 will need to be capable of carrying them all.

For now it’s more a question of plugging the gaps where necessary, for instance by developing intelligent gateways that can collect and analyse data before passing it on into the cloud. A kind of way of collecting together the rain drops, you could say. Those kind of issues are increasingly being addressed by IT companies.

Another design opportunity in the IoT of today – which poses an entirely different and, one could say, more philosophical question – is how does one start to approach the question of device and data security, when there are devices from myriad different manufacturers, a range of protocols being used and many different types of network architecture are involved?  There’s recently been an admission from the US Director of National Intelligence, James Clapper that, in the future, the IoT might be used by intelligence services “for identification, surveillance, monitoring, location tracking, and targeting for recruitment”.

When Dick Cheney was in office, his cardiologist reputedly turned off his pacemaker’s WiFi capability just in case the device was hacked – and US students last year succeeded in hacking into the pacemaker of a robotic dummy used to train medical students and effectively delivered the coup de grâce. Think then about other medical devices that are controlled to deliver life-preserving doses of drugs such as insulin, or are hooked up to the legacy systems with which health care systems are liberally populated and that run outdated operating systems or have software that’s so old it can’t be updated.

Late last year, research company Forrester predicted that “2016 will be the year we see ransomware for a medical device or wearable”.

This is scary stuff for a concept that seems, on the surface, to offer so much to so many. But it’s a problem that has to be thought about at every stage of the design process. It’s not really enough trying to engage in a massive retrofit operation – although that is undoubtedly going to have to happen (and quickly) for some of the extant devices. Security and the dangers posed by hackers need to be uppermost in the mind of every design engineer and circuit designer. Out there, someone’s life could depend on it.

That’s throwing down the gauntlet. Are there strategies you could adopt to get round the problem? How can designers influence the development and future direction of the IoT?

Or do you think this is overblowing the issue? Is there really a danger posed by hackers?

Oh Brave New World!