At MWC this year, Canonical and Vodafone announced a partnership to test a new, cloud-driven “smartphone”. Vodafone will use Canonical’s Anbox Cloud to host its Android operating system, which would be able to run – and perform similarly to a smartphone – on less complicated computing devices, TVs, wearables and more.

The experience of running an operating system, its apps and services would be handed over to the cloud-based service, while the end device simply serves as the interface.

On the surface this seems like a sensible idea. Smartphones can be enormously compute-intensive, creating problems for battery life. Offloading intelligence and computing can also lead to significantly cheaper devices, improving the affordability of services and extending their reach to lower-income countries and groups globally.

Thirdly – it means that the end device is not, in large part, responsible for the performance the customer receives. As a result, it would reduce the demand for device upgrades, extending average device lifetimes and reducing the environmental pressures caused by today’s smartphone lifecycle. 

Questions, Questions

The concept of offloading the compute and storage, using only a “thin client” on a phone, has been around for over a decade. There have been good reasons why they have not become mainstream yet.    

Firstly, and most significantly, thin clients depend not only on the server but also on the ability to connect to it with sufficient quality and consistency. While in many countries fixed broadband is the “gold standard” that 5G is aiming to rival, many of us will sympathise with the frustrations of video calls dropping out or breaking up. Consistently good quality coverage would be necessary for a thin client solution to work. Arguably not a problem for a TV in a home served by fibre broadband, but risky as a mobile proposition.

This restriction is also applicable in countries, and in areas within countries, with less developed networks – frequently the kinds of places that would benefit most from low-cost, thin-client “smartphones”.

One more consideration could place strains on the model even in areas well served with fibre networks. As Sue Rudd, Director of Networks and Service Platforms at Strategy Analytics, pointed out in this recent webinar, 5G itself might “break the internet”.

“We’ve got a model today where the access is faster than the backbone of the internet. 20,000 people in a stadium sending half a gig is faster than the 50G or 100G backhaul,” she observed.

If the on-device computing moved to the cloud it would effectively make all traffic flows to devices similar to streaming video, sending much more traffic back and forth from the core. Operators would be faced with massive extra investments in fibre or poor service delivery.

Historically Content Delivery Networks (CDNs) have at least in part helped to reduce the core network burden by caching content in multiple locations closer to end users. A similar solution using edge compute and storage nodes could reduce the dependency on the backbone network.

Vodafone launched Multi-Access Edge Compute (MEC) solutions for business in the UK and Germany in 2021, so it’s not far-fetched to think that this combination of edge and thin clients might be deployed more widely.

Meanwhile, Vodafone’s African unit Vodacom has been rolling out 5G for fixed wireless access and mobile broadband since 2020, including in 700MHz spectrum for wide area coverage. While it is far from universal in South Africa, let alone elsewhere on the continent, there is perhaps an argument that this could be used to connect some consumer devices with thin clients for local use; particularly if MEC is also rolled out to support local requirements.

One strategic pillar of Vodafone Group’s overall strategy is to become an “African Technology Leader” through its Vodacom operating companies. Aggressive moves to deploy 5G more widely would enable Vodacom to offer consumers and businesses advanced services, and types of service, in ways that would deliver on that strategic aim. A thin-client, cloud-based service would certainly enable distinctive services provided it had the network to back it up.

Getting Personal

Meanwhile, the concept of creating a cellular “Personal-Area Network” (PAN) already exists, especially using Wi-Fi to connect devices in a small area around the home.

This model of personal area networking is quite different to what Vodafone is suggesting, however.

Currently, TVs and other major devices tend to have their own connections through a Wi-Fi router, and their own intelligence so that they can be sold individually. Networking devices together is only possible in some circumstances. The closest we come to Vodafone’s vision are some very specific and closed ecosystems such as Apple’s, or smart home systems such as Hive.

A Vodafone-style model with thin clients could – in theory – replace some of the more expensive smart devices in the home. However, there is arguably a better commercial opportunity to provide the “brains” of less expensive devices that would benefit from a measure of intelligence, along the lines of Hive’s services or wearables. Using a cloud-based instance of Android as the unifying operating system across these devices would be an appealing prospect.

It would all depend, though, on having a connection back to a storage and computing asset that can be relied upon. In this case, the fewer hops to the end device, the less risk of a break in service. A good smartphone might be the storage and computing asset in a PAN, assuming that the demands from the other devices aren’t too great. Otherwise, a home router might be made to serve a similar function, or at worst connect over fibre or a similar 5G backhaul link to the network edge. As mentioned, though, the more hops the greater risk of breaks in service.

Beyond Personal

The one area where thin clients are more commonly used are in office situations, with central servers supporting thin-client devices which act effectively as desktop computers or similar workstations. The advantage here is clearly that the company owning the office has more control over security, data management and more.

The pandemic has seen a major shift away from offices, and there is a tendency for commentators to think that we will not see people in offices to the extent we previously did. There is – perhaps – an opportunity to help companies to adapt their current equipment for an environment with increased homeworking.

Louder Than Words

Overall, Vodafone’s demonstration at MWC does not necessarily demonstrate current capabilities or even a business plan, but it does illustrate some intentions; in this case, to go beyond being a connectivity provider and becoming more involved as a platform for device management and services.

Given Vodafone’s fixed assets there are ways for the company to start playing with thin clients in the enterprise space reasonably soon, but if they are serious about delivering “smartphones” without the smartness then much depends upon the evolution of their edge and 5G coverage and capacity.

There are clearly a variety of business models that could grow from this; the big question will be what Vodafone focuses on, in what markets… and on what timescale. Having deployed their first 5G site using Open RAN in January, Vodafone aims to have 30% of its sites in Europe using it by 2030. Let’s hope that their plans to support thin clients can scale faster.