With 5G, mature applications are destined to be dumb
With 5G on the horizon, the promise of hyper-fast connection speeds is near. By making network communication delays — or latency — largely a thing of the past, 5G will change the way companies manage their workloads. For many IoT, browser and mobile applications, business logic, computational power and data storage will move to the cloud and data centers, transforming devices from handheld computers into simple UI layers — or dumb devices.
This trend of offloading business logic from a device to distributed cloud locations will no doubt grow IoT, which has virtually no storage limitations. Back ends today can store terabytes or petabytes of data versus mere gigabytes on mobile phones, and offloading means we’re no longer limited to data from local databases.
There’s an ongoing debate on the merits of processing and analyzing data on the edge versus in the cloud or back end. But the distinction is simple. IIoT, specifically environments in rugged and remote regions, will rely more on edge computing, which takes on critical importance in extreme environments — think airplanes in flight or sensors on Arctic icebergs — where you can’t rely on anything but the device itself for decision-making. But non-IIoT mobile and browser applications will increasingly offload computation to the cloud as 5G takes hold.
It’s equally important to draw a distinction between processing trends for mature applications, such as mobile and browser apps, and those running in an industrial environment or autonomous vehicle, where a delay in decision-making could harm humans. In the section below, I focus exclusively on 5G’s impact on mature applications, and will save a broader discussion of edge computing for a future post.
In short, migrating a device’s brains to the cloud will come with many technical and business challenges. Here are the top three concerns and how companies can prepare for 5G.
Concern #1: How will 5G impact customers?
Key performance indicators will need to change to take lower latency into account. For example, in a 5G world, should your business consider time spent on application a success metric? After all, faster connections will enable your customers to achieve their goals faster, and less time spent with an app may not necessarily mean lower engagement.
Instead, it might be wiser to focus on the combined impact of revenue, geography and connection speed. Is there, for instance, a significant difference in the customer experience for users who access your service via 5G versus 4G?
Concern #2: How will 5G impact resources?
As complexity migrates from end devices to the cloud, the front-end mature application will grow dumber, essentially becoming a UI layer with the business logic living in the cloud. This dramatic change will require significantly more computational power and storage at the cloud level, and businesses will need new ways to glean real-time insights from the surging volume, velocity and variety of data. Be on the lookout for ever growing cloud computing bills form AWS, Azure or Google Cloud Platform.
To manage this complexity, your business will need tighter integration between network and application to see exactly where problems arise. We’re already seeing this trend with streaming analytics: Lacking sufficient time to store incoming data, companies are extracting insights as the data flows in. This process will grow more complex as faster 5G networks allow businesses to ingest even more information.
Concern #3: What development platforms and measurement tools will enterprises use on the front end?
If the front end does turn primarily into a UI layer, enterprises will be able to move across different development platforms more easily, relatively speaking. We are already seeing the proliferation of new mobile app development languages, such as React Native from Facebook, Xamarin from Microsoft and Flutter from Google.
Measurement tools will need to re-innovate to embrace the emerging AIOps trend, where businesses use artificial intelligence, machine learning, big data and other powerful analytics technologies to enhance ITOps functions with proactive insight. Performance alerts will need to be more accurate. Systems will need to detect and track customer journeys without human input. And, in the near future, AIOps-based self-healing capabilities will need to be built into vendor tools.
Prepping for a 5G world
5G is more than an evolutionary upgrade of 4G. It’s a revolutionary advancement in communications that will enable everything from remote surgery to smart connected cities. We’re already catching a glimpse of the advanced applications 5G will enable. At MWC19 in Barcelona, for instance, Ericsson demoed a new teleoperation system where a driver could remotely control an autonomous, electric truck at a test site more than 2,000 kilometers away in Sweden.
To take advantage of 5G, businesses will need enhanced monitoring, analytics and automation tools designed for complex cloud environments that do much of the heavy lifting. The front-end mature app? It’s destined to be dumb.
All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.