Adam de Linde from Orange showed the Mobile UX London conference how understanding social organisation and work practices matter when designing what he calls ‘5G environments’, the convergence of 5G networks, edge computing and AI.
On November 21st 2019, the 2019 edition of Mobile UX London took place at the DeVere West One in London’s West End. The audience of mobile UX practitioners heard from a world-class line-up of speakers, and took part in a wide range of practical workshops. It was a great day and I know everyone who attended took value from it.
We’ve put together a series of articles to recap on some of the presentations, so even if you couldn’t be there, you can still take in what happened.
Adam de Linde
It was a pleasure to have Adam de Linde speak at MUXL. Adam is Design Director at Orange, one of the world’s telco giants. His role includes promoting service design for large-scale IoT projects.
The title of Adam’s talk was ‘Designing with 5G environments’. In his talk, Adam explored some of the emerging characteristics of ‘smart networks’ and their effect on work environments. He also proposed some ways to frame design challenges when considering environments with intelligent infrastructure.
What are 5G environments?
Here are some of the highlights of Adam’s presentation.
Before explaining 5G environments, it’s helpful to clarify what 5G itself is. The 5G standard offers enhanced mobile broadband, meaning giga-bytes per second (Gbps) data rates, ultra-reliability and low latency, in the millisecond range, and support for massive machine type communication, around a million connected devices within a square kilometre area.
5G combines a host of transformative technologies for networks. However this transformation achieves still greater significance when coupled with other parallel developments:
IoT – It’s estimated there will be in the region of 50 billion connected devices by 2023 and 75 billion by 2025; everything from cars to cameras, VR headsets and factory sensors, all generating quantities of data that would saturate the networks
Edge computing – processing information closer to the location of the device, away from the data center, achieves both quicker responses for the devices and reduces the amount of traffic going back and forth to data centres.
AI and machine learning – typically the type of computing performed at the edge is of the AI/ML variety. Making inferences at the edge, such as image recognition, saves enormous time and bandwidth compared with transferring raw data back to data-centres for processing.
It’s really this combination of 5G capabilities with intelligence at the edge that creates what Adam calls ‘5G environments’. The term is loosely interchangeable with ‘ubiquitous computing’ or ‘ambient connectivity’, however referring to ‘environments’ emphasises the specific user context and their role within it.
The potential of this technology is unparalleled, but Adam showed us some examples where 5G environments are already making a difference.
Real-world examples of 5G environments – The shipyard
Adam’s first example of 5G environments in today’s world of work was a shipyard in Northern France, where they build massive cruise ships. Although it started as a technically-led project, Adam’s design team approached this with a holistic ‘service design’ mind-set.
A shipyard is a big, complex, hazardous, noisy place, with a lot of people and lot of equipment. The owners of the shipyard wanted a way to monitor the equipment because a lot of it was going missing; these sophisticated tools cost a lot to replace, so anything that could reduce these losses would be desirable.
However, the shipyard didn’t want to say they were installing monitoring technology because their staff can’t be trusted, so the initiative needed to be reframed. Instead, Adam and his team decided to look at the work process, how they allocate tasks and how workers find the equipment they need. Monitoring equipment would be secondary.
One part of this challenge was technical. Because large ships are made of steel, they are in effect giant Faraday cages, blocking out radio signals. This was overcome by installing beacons within the ship’s structure for the internal tracking, with relays to the external network for continuous tracking across the entire site.
The beacons allowed tracking of tagged materials and equipment from storage through to installation in the construction. But they were found to have other benefits as well. Since the beacons could be easily repositioned as construction of the ship progressed, they could direct workers via optimal routes to the location of their allocated tasks, which offers significant time savings considering these vessels are hundreds of meters long and several storeys high.
The beacons also offered a safety benefit, since they could be used to send a notification in the event of accident or injury, with a precise location marker.
Providing this level of connectivity generates what is known as a ‘Digital Twin’, essentially a virtual model of the activity involved in the ship’s construction. This then becomes a powerful tool for optimising processes and workflows, coming back to the point that human workflows and organisation are really the vital elements that need to be understood and modelled in any design approach.
Other examples of 5G environments
5G and IoT are massively important in the industrial and manufacturing sectors because of the productivity gains they generate. But Adam showed that 5G can also apply to consumer settings. Perhaps not directly via 5G handsets but through ‘smart environments’ that can understand and interpret consumer behaviours through a combination of edge intelligence and 5G.
An obvious application of this is the analysis of video footage from the millions of cameras that track our daily movements, whether in urban settings, on transport networks, or retail spaces.
A couple of examples served to highlight this: A Sao Paolo metro line that uses AI-enabled cameras to profile passengers waiting on platforms to board their trains, in order to show targeted adverts on large media display screens nearby.
There was also the example of the UK shopping centre using cameras to track consumers around the mall, with the ability to detect attributes such as age, gender and time spent in stores in order to provide more accurate profiling to retailers in real-time – effectively giving retailers the ability to use contextual data to tailor customer service in their physical stores.
These are not far-fetched realities. These are services already being experimented with today, and in Adam’s opinion have a pressing need for designers to engage with questions of inference from data with the goal to provide meaningful and fair-use services.
A more light-hearted example was provided by the AI-Bar in Shoreditch, London. Facial recognition tracks the arrival of clients at the bar, placing them in a ‘virtual queue’ displayed on a screen over the bar which ensures bar staff can serve clients according to their order of arrival. A sort of digital-twin of the bar, if you will.
Adam highlighted the need for transparency in respect to usage of the data, contrasting the AI-Bar, where the information and its purpose is available for all to see, with the shopping mall where inferences about consumer behaviour directly observed from the video streams are exploitable by the retailers, leading to arguably better service but without fully recognising the users’ choices and preferences.
The need for balance between service value and the exploitation of data will always have to be negotiated, but it’s where a design standpoint can provide a mediating role.
This was demonstrated through his final example of the ‘smart crossing’, an AI-enabled pedestrian-crossing in London. This was a design intervention by London studio Mettle that firstly shows you don’t have to be a big telco to ‘hack’ intelligent environments. Secondly it emphasised the importance of the context of use, not just the technology.
The smart-crossing uses AI-enabled cameras at a road crossing to detect the approach of pedestrians and signal to approaching vehicles via a light-strip embedded into the road. However, rather than looking at the problem from the point of view of collision detection, their installation served to explore the social context of the crossing – raising question about pedestrians with varied levels and means of mobility, or the usage of the street at different times of day, for example when kids crowd the pavements going to or from school.
The example served to support his concluding remarks, that while deploying 5G and intelligent edge infrastructure we also need to design for context, factoring in social activity as part of the design approach, or in the case of industrial settings the work practices themselves as part of the infrastructural design. With contextual awareness comes the recognition that users act not only as consumers of technology and services, they are active contributors in forming and sustaining such environments. Designers have a key role to play here, going beyond user-centred design to articulate collective value in digitalised environments.
Stay in touch
Thanks to Adam for sharing his insights with the audience. It was a thought-provoking talk, which I know everyone in attendance enjoyed.
Over the next few days and weeks, we’ll be sharing articles and videos from the event, so make sure you stay in touch with Mobile UX London. Subscribe to our newsletter to make sure you never miss a thing.