IoT has been a buzzword for almost two decades. Are we finally at the point where we can leverage its potential in everyday life? With increased adoption of advanced computational capabilities, like machine learning, we are finally starting to see IoT maturing into IoE ("Internet of Everything"). What might that mean for healthcare?
We first half-facetiously described the Internet of Things (or IoT) as “The Internet of Toasters” back in November – based on an example described by Justin Brookman, policy director of the FTC’s new Office of Technology Research and Investigation. He represented the “Thing” in IoT as devices or objects that perform a certain task (like a toaster), and at the same time, collect data about how/when those tasks are performed. The ability to connect and share this data across multiple devices with the goal to optimize an individual’s life is the ultimate vision behind IoT. But are we getting any closer to realizing this potential, or are we still just connecting a bunch of toasters?
With the growth of connected devices (99% of electronics currently not connected will be connected in the next decade), enabled by recent trends in artificial intelligence, we are now witnessing the evolution of IoT to IoE (the “Internet of Everything”) – where autonomous devices, ubiquitous connectivity, and intelligent computing are setting the stage to truly deliver on “living services.”
In this post, we will explore the common barriers to IoE, how they will be overcome with advanced computational capabilities, and specific applications of IoE in healthcare (where it is believed to be the fastest growing industry).
Breaking Down Barriers to IoE
There are already big problems to solve with computers and devices at the current state today, but all these problems are amplified when these devices are connected and independent. Specifically, some of the biggest challenges with IoE are Security, Privacy and Reliability.
Security and Reliability are mainly technical challenges that we have been dealing with long before the internet, from the famous Caesar cipher, to Enigma, to modern day encryption, and zero day exploits. Information Security is the practice of defending information from unauthorized access or any other threats. It’s a unique challenge, that currently requires good information security hygiene to provide around 80% security.
However, IoE amplifies security risks exponentially; unsupervised and self-aware IoE is in desperate need of a whole re-think of security, reliability, and how we deal with complexity. Perhaps it’s finally time to admit that only Artificial Intelligence and machines that can learn and adapt are flawless enough to deal with the problem at such scale. Stay tuned for a follow up post that dives deeper into this topic.
Privacy is another major concern which is in part related to technology, and in part related to how we operate as a society. The massive amount of (mainly) personal data generated brings massive challenges and can cause profound political, social, and economic transformations. It is essential that users are always in power of their own data, and everything surrounding them behaves with adequate accountability and transparency.
Protocols that make everything self-aware, connectable, and allow for different data flows are the way forward to allow for peer to peer independence, and let users be in full control.
In addition to the technical challenges, the other challenge to adoption is related to usability.
Connected devices by definition can, and will generate massive amounts of raw data. However, there hasn’t been an efficient and meaningful way to use this data. Properly contextualized data that can deliver actionable business insights is estimated to be only around 15% of the total amount of data generated by IoE. Finding insights in terabytes of machine data, especially when the data is as raw and unstructured as it is, will simply not be resolved by analytics and stats alone. It is unreasonable to expect humans and technologies that do not scale, to contextualize more than (what we actually can analyze is even less than 15%).
Machine learning can be used to run through raw data, structures the data, learns from it, finds suggestions, interprets, and allows for actionable insights. These insights can finally be used to provide a more optimal user experience – which has always been a barrier to maximizing the potential of IoE. While smart devices were always meant to make our lives easier, users typically don’t know how to make sense of this data or what to do as a result of seeing their data, and therefore give up using the device altogether. Enabled by intelligence, we are finally able to move from a device-centric ecosystem to a truly human-centric ecosystem.
IoE Adoption in Health
Perhaps the most talked about “Thing” as it relates to healthcare has been wearable devices. While these wearables have come a long way in terms of adoption (as of March, 2016, over 10% of US consumers now own a wearable device, with FitBit capturing the majority of the market), they are only truly impacting health outcomes if the data that is collected can be connected, via an IoT platform, and can be leveraged by other devices, shared with doctors or other stakeholders, in a timely and relevant manner.
Therefore, as discussed, a key part of enabling IoE will be the ability to leverage machine learning capabilities to truly deliver a human-centric experience. While we haven’t seen this happening at scale yet, new market entrants show promise, leveraging the power of IoE to increase speed and accuracy of diagnosis and access to care.
One such product is ResApp, a smartphone app used to help patients diagnose and monitor a range of respiratory diseases right on their phone. The app functions by capturing patient’s breathing patterns on their phone, and applying a machine learning algorithm to most accurately determine which type of respiratory disease, as well as severity. This data can then be shared with doctors, allowing doctors to treat patients remotely.
Another application is in disease management and support. One of the big problems in healthcare is adherence, and solutions such as AdhereTech’s smart wireless pill bottles are already being used to collect and send adherence data in real-time. The data gets analyzed and customizable alerts or interventions can be sent using a range of communication channels. Data analysis gets better as more users are using it, and more insights are gained.
The next natural step would be to accurately determine the reasons for non-compliance. These reasons can be practical, physiological, emotional or personal. In the future, by integrating AdhereTech with IoE, we can get a more holistic picture of patients’ experiences with their treatment.
Using frameworks like CareKit to integrate everything together, we can combine data from multiple devices and sensors: activity trackers, symptom trackers and other available data sources. Layering machine learning models, we can even start to predict when patients may stop taking medications, finally determine the true reasons, and use this knowledge to take meaningful action to improve adherence.
Perhaps the most comprehensive use of IoE in health we have seen to date is a study being conducted by Pfizer and IBM, using a “smart” home filled with sensors to continuously track, in real time, data from Parkinson’s patients. Machine learning can then be applied to determine the impact of medication on various “quality of life” aspects. Through applications like this, IoE is not only delivering better care, but also helping to enable more accurate research leading to better treatment.
Look to the Future
A future in which we all live in connected homes, where every device is digitally enabled and has the ability to talk to each other, is perhaps not as far off as we thought. We will essentially live in a connected digital ecosystem, where a continuous stream of data is being collected about us in real time, data which not only provides insights into our mood, activity level, sleep patterns, but also our health and well being. There will be an emergence of new data types and derived data types (e.g. think of google traffic data, which is generated by collecting the speed of moving cars from crowdsourced user generated data from mobile phones). These devices will become increasingly “aware” of other devices that are around them, and smart enough to know what device, when, and to whom to send this data – healthcare providers, caregivers, and other stakeholders – all with the goal of optimizing our lives.
How can we as healthcare marketers ultimately leverage this ecosystem to deliver better care to patients?
One can imagine a scenario where a fridge could provide food shopping and recipe recommendations based on an individual’s diet and exercise data generated from their tracking device. When medication runs low, their pill bottle can send an alert to the pharmacy where a drone will automatically deliver medication to patient’s doorstep. Alerts and recommendations will be completely personalized based on how a patient engages with their health and treatment journey.
While creating the right patient experience in the world of IoE may seem complex, leveraging the right platform enabled with the right intelligence can greatly help accelerate the adoption process.
Here at Klick, we are committed to delivering on the IoE experience for all healthcare stakeholders to drive better health outcomes. Our team of digital explorers and inventors use an omni-channel approach to bring personally resonant stories of treatment and recovery to life, through a variety of mediums, enabled by intelligent platforms and seamless device connectivity, all while ensuring the highest degree of data security and privacy with our team of leading security experts.
Want to learn more about how you can take part in IoE revolution? Reach out to Klick LABS, our very own dedicated digital innovation team, to start the conversation!