Over the course of the last 15 years, Transforma Insights’ analysts have tracked the growing use of remote monitoring and management of a diverse array of distributed assets, including cars, industrial equipment, energy infrastructure and healthcare devices, deployed for the purposes of improving customer experience, reducing maintenance costs and compliance with regulation, amongst other things. Collectively this trend has been termed the Internet of Things (IoT), writes Matt Hatton.

For the most part, IoT devices have been used for data transfer. However, there is a significant sub-set of IoT devices that also need to support voice services of various types. Some IoT use cases are enhanced by two-way communications (such as for doorbells), whereas in others there are regulatory mandates requiring the ability to initiate calls, for instance to emergency services (such as in eCall).

In this article we explore the growing requirement to support voice services in IoT and the growing requirement for a scalable VoLTE/VoNR/IMS capability to support it.

Many critical IoT applications require voice services

Cellular IoT applications demanding voice services can be broadly split into two main categories: automotive and non-automotive, reflecting the importance of the auto sector in cellular IoT.

Collectively, according to Transforma Insights’ IoT Forecast Database, connected vehicles accounts for 33% of cellular-based IoT connections and 46% of cellular connectivity revenue, as of 2023. As such, the requirement for supporting voice within this sub-set of IoT will be a major consideration. Numerous automotive applications need to support some form of voice service, whether it be for automated emergency call (eCall), roadside assistance or fleet management dispatch, amongst others.

In addition, there is a wide variety of other use cases that also demand support for voice, including intercoms, lone worker safety, connected elevators and assisted living.

In total, 22% of the IoT connectivity opportunity will require voice support

Based on Transforma Insights’ highly granular IoT market forecasts we can identify the share of IoT connections that will have a requirement for voice services. By 2033 almost 20% of IoT devices are in a category of use case that has a requirement for voice support, up from 14% in 2023. The 2024 figure equates to 1.4 billion connections.

Turning to the revenue associated with these IoT connections, the proportion of revenue that is dependent on supporting voice services is even more significant over the forecast period. By 2033 of the almost US$80 billion global cellular IoT connectivity revenue opportunity, 22% is generated by applications that will have a requirement for voice support. Over 60% of that opportunity is accounted for by automotive applications.

Supporting VoLTE and eventually VoNR is critical

There are many forms of voice communication covered by the diverse range of use cases that Transforma Insights has identified. What is mostly common amongst them is a requirement to support native IP voice in the form of Voice-over-LTE (VoLTE) today, and eventually Voice over New Radio (VoNR) as 5G becomes more pervasive. There are several reasons for this.

With the switch-off of 2G and 3G networks, the ability to use circuit switched (CS) calling for voice will disappear. Until now, even devices predominantly using 4G connectivity have had the option of circuit switched fall back (CSFB) whereby a multi-mode device would use 2G/3G networks for the voice call element of the application. With 2G/3G refarming, that option will be increasingly rare.

Furthermore, many IoT applications are for what we might term ‘critical’ applications, where the use of robust and secure mechanisms for communication will be paramount. In some cases the technology to be used will even be specified by regulations, as is the case with eCall. In addition, many have a requirement to integrate secondary information with the call, such as video feeds, patient data, or location. With that being the case it will be largely inappropriate for voice services to be provided using alternative approaches, in the form of over-the-top (OTT) voice. This generally lacks the ability to integrate voice with other data as part of a session as we see with IP multimedia system (IMS), one of the technology underpinnings of VoLTE.

We should note that for some use cases, OTT voice will be perfectly adequate, but in the majority of use cases it will not.

The result: IoT connectivity providers need a scalable VoLTE/VoNR solution

An increasing amount of the IoT connectivity market depends on supporting voice services. As such any communications service providers (CSPs) delivering IoT connectivity will need to be able to deliver voice services, and specifically VoLTE. With revenue per connection continuing to decline, they will need a cost-effective and scalable approach to supporting it.

The requirement for delivering voice services causes challenges for many IoT CSPs. Most are dependent on inefficient legacy solutions that are not future-proof and/or scalable and – most importantly – cannot be delivered at the type of cost structures IoT customers are expecting.

Transforma Insights notes eleven characteristics of an optimised voice capability for IoT. This ranges from being secure and compliant to being flexible in call routing and able to enrich calls with contextual information. IoT connectivity providers should demand that any voice solution they deploy meets these requirements.

CSPs must give careful consideration to scalability of the voice technology solution. Few will have millions of dollars spare to invest in over-provisioned functionality. Instead, they will want a minimal up-front commitment and then scale by number of users with proven technology scalability.