IoT & Realtime Communications

Dean Bubley
March 8, 2016


We have grown used to the idea of devices and machines communicating between themselves, and with IT systems. From the early days of machine-to-machine connectivity, to today's expansive visions for IoT, we expect networked sensors, actuators, consumer electronics and other devices to send data in both realtime and in delayed/batched messages.

But as well as the "things" and their related back-end systems, many IoT implementations also involve communication with one or more humans either directly or indirectly. While some instances are relatively mundane – for example, devices sending notifications or text-based messages – a subset involve realtime communications; that is, voice and video. This could just be an IoT event triggering a traditional phone call, but more interesting and unconventional variants are appearing as well.

For example, if a sensor records excess temperatures in a machine, a remote supervisor may need to speak to a technician, or use a video-camera to check the situation visually for signs of smoke or fire, or perhaps an audio circuit to listen for unusual mechanical sounds. Other scenarios will apply to different sectors – perhaps an agricultural drone being flown by pilot from a tablet, speaking to a farmer or specialist at the same time, about crop conditions.

At first sight this might appear mundane – essentially an IT system setting up a phone call between relevant parties, perhaps with simple integration with the PBX, videoconference system or UC system in a business.

Communications technology trends

But this overlooks trends in the communications technology sphere, which is itself evolving in numerous ways in parallel with IoT's development. For example:

Mobile is obviously a critical trend – much of the time, a given person involved in an IoT-triggered session will be using a smartphone or tablet, and will tend to use native apps, rather than dialler or web browser.

Embedded communications, in which voice, video or data streams are integrated directly into an application or web site, rather than as standalone capabilities. For example, a field-maintenance app may incorporate "see what I see" video-conferencing between the local engineer and a specialist called in to help diagnose a problem.

Contextual communications, in which the human voice/video functions are linked to analytical observations, sensor input or event. This could relate to machine-learned preferences for the people involved, or local sensors such as a smartphone's motion-detection.

Workstreams are becoming an important metaphor in communications. Familiar to many people from social-networking, we are seeing wide adoption of "scrollable timelines" of messages, voice/video sessions, statuses and integrations from 3rd-party apps and devices. A city administrator may get feeds from various systems covering transport, waste-management, environment, power usage and so on, fed into a dashboard or workstream display, and be able to share or discuss items with colleagues in realtime.

As well as IoT events acting as triggers for voice or video sessions, sometimes the chain of causality will be reversed. A normal "human" communications session may become enhanced through the incorporation of additional IoT connectivity – for example, a doctor undertaking a telemedicine consulting session with a remote patient may blend in streamed data, perhaps from a blood-pressure sensor or an ultrasound detector, connected to the same PC or phone the patient is using to speak.

Thus we will see IoT interactions incorporate RTC (realtime communications), as well as RTC sessions incorporating IoT. In other words, the number of possible "touch points" between IoT and RTC domains is proliferating, and the trajectory of the technologies in both areas suggests further coupling in future. Clearly, only a fraction of IoT operations will involve voice or video connections – but in a world of tens of billions of devices, even an occasional input or action will drive a sizeable addition to the pantheon of human communications. It is much less clear, however, whether the traditional telecoms industry is going to be in a position to benefit from this, given the fragmentation and complex value chains involved. Value will be driven by outcomes and productivity – not by minutes of voice or video.

An important point to note is that many of these instances of RTC will not resemble traditional "phone calls" – i.e. two people interacting for a prolonged period, using conventional identifiers (phone numbers) and standardised call-signalling or metaphors such as "ringing". Nor will the video sessions always be "talking heads" like a FaceTime or Skype call. We will see a mix of two-way, one-way, synchronous, multi-party, one-way only, front vs. back cameras, and alternative uses of audio and video such as machine-optimised microphones or infrared cameras.

Consider a public-safety use-case: applications might blend hands-free walkie-talkies, body cameras, gunshot detection, heat sensors, video feeds from drones or dash-cams, realtime language-interpretation or lie-detection and numerous other manifestations of RTC. These blend new devices for input and output, plus a mix of human and machine communications.

Often, these IoT-driven human interactions will involve mobile apps or web-browsers. From a technology standpoint, this requires the use of tools and platforms that allow developers to create a broad diversity of interaction models, span a broad range of connectivity options, sometimes apply extra quality/performance techniques, and sometimes use media-processing capabilities on the realtime feed. In many instances, it will be designers rather than engineers that take the lead on service creation.

IP-based mashups

While some instances will integrate with traditional telephony or SMS messaging, the vast bulk of IoT/RTC mashups will be purely IP-based. Most will work outside the traditional telecoms world of IMS and VoLTE as well, which are insufficiently flexible to accommodate the range of use-cases highlighted here. Various new developer-friendly RTC approaches will be used instead, facilitating cloud- and premise-based solutions for developers.

While there are various proprietary tools and cloud-based PaaS (platforms-as-a-service) designed to embed RTC into applications, the most important emerging standard is WebRTC and its close relation ORTC (Object RTC). Already supported in the bulk of browsers, and easily embedded into mobile apps on most smartphones, it is already powering many of the more recent video- and voice-based communications applications, as well as embedded features in many websites. A number of vendors are already pitching IoT integration capabilities in gateways and middleware, intended for a variety of implementation styles. In 2016 and 2017, these are likely to become increasingly well-featured and reliable.

In summary, the worlds of IoT and RTC are intersecting, and the overlap is likely to widen significantly in 2016 and 2017. Perhaps the "killer use-case" is around preventative maintenance for machinery – if malfunctioning systems can be pre-emptively diagnosed and fixed before they break rather than afterwards, there is a huge upside for companies and economies. Yet such interventions will almost always need human communications – promptly, and with detailed contextual information. Healthcare is another promising domain, linking physicians, patients and the multiplicity of networked equipment in rapid and secure fashion.

We are still only at the early stages of IoT/RTC integration, but it seems as though the provision of the requisite gateways, client/device integration and middleware systems holds significant opportunity, as does the design and realisation of new hybrid solutions.



Dean BubleyDean Bubley is the Founder of Disruptive Analysis, an independent technology industry analyst and consulting firm. An analyst with over 20 years' experience, he specialises in wireless, mobile, and telecoms fields. He is one of the leading analysts covering Mobile Networks, Service Provider business models, The Future of Voice and WebRTC.

As @disruptivedean on Twitter and through articles on the Disruptive Wireless blog and LinkedIn, he gives informed, contrarian commentary on the telecoms and IT industries, as well as policy and regulation. He speaks at 30+ events per year and is regularly quoted in the media He also works collaboratively with a number of other leading experts and consultancies.