You don’t need me to tell you that we are living in the data-driven age. Consumers and businesses alike are increasingly using digital for everything, IoT devices continue to proliferate and as 5G and machine-to-machine communication is adopted, there will be even more data to deal with.
The shift towards cloud has helped enterprises make their data perform and deliver business value, as well as raising the bar in terms of business agility, speed and performance. However, as enterprises process more data in real time, they need to address the latency issues inherent in sending all that data to the cloud and then back again.
While a little bit of lag time is inconvenient and detracts from the user experience in some scenarios, latency can be downright dangerous in others: think medical devices, emergency services or self-driving vehicles.
The answer? Process more of the data closer to the user, aka edge computing.
How Edge Computing Works
Edge computing refers to data that is processed at or near its source. That could be near the user, for example, a phone, smartwatch or medical monitoring device, or it can be near another source, like an embedded sensor in any kind of internet-enabled device.
Edge computing is growing fast. By 2025, Gartner predicts that three quarters of data generated will be processed outside a traditional centralized data center or cloud, up from 10 percent in 2018.
While it is not new per se, edge computing is becoming an increasingly important consideration for businesses embarking on some of the hottest growth areas and megatrends in digital right now – these include (almost inevitably) IoT, AI, digital payments, medical devices, connected cars and more. And, because of continued investment, edge computing solutions represent innovative opportunities to improve the performance of existing apps as well as representing significant growth potential for new use cases.
Not all Edges are Equal
The caveat to this increased awareness is that when enterprises start to explore the edge computing options, things can begin to get confusing. On a very simple level, the right solution for an initiative depends on its specific requirements and how they map on to the company’s existing vendor relationships, infrastructure, skills and resources.
Furthermore, there are differences in the way vendors, analysts and the industry at large talk about edge computing.
For example, Forrester’s recent report – “The Four Edges of Edge Computing” – identifies four different types while the analyst also notes that different vendors use the term to refer to different aspects of edge computing. When you think about how much more awareness has been generated in recent months, it is therefore important for enterprises to understand that comparisons between solution providers may not be like-for-like.
How should Enterprises Approach Edge Computing Solutions?
In September of last year, IDC’s Dave McCarthy described edge products and services as “powering the next wave of digital transformation.”
With so many solutions and vendors to choose from, enterprises will naturally want to research the edge landscape before investing. Given that IDC predicts the edge computing market will show a 12.5 percent CAGR through 2024, companies will also want to future-proof their investments.
A good, reliable place to start is AWS’s service offering, AWS for the Edge, which spans everything from edge/cloud infrastructure and storage services through to content delivery, robotics, ML, IoT and solutions for rugged conditions when network connectivity is unreliable.
Experience at the Edge
When businesses embark on a new edge deployment, creating a list of core business requirements for the solution should be a priority. And yet, even assuming there’s a clear vision of what the enterprise’s goals are, selecting between specialist edge vendors and the large-scale cloud infrastructure offerings available can be difficult.
We may be biased but given that Infostretch has worked on IoT initiatives with the largest chipset manufacturers right through to early-stage disruptors in industries as varied as healthcare and connected cars, we would advise asking your trusted digital engineering partner for guidance. And a digital engineering services firm – like Infostretch – should have a lot of experience to draw from.
In the IoT space, for example, expect to be shown a proven track record of converting analog-based use cases to digital using sensors, IoT chipsets, firmware development and creating the full-stack infrastructure around it – web, mobile, cloud and analytics. As well as our IoT Development Services, Infostretch created a remote device management platform especially developed for IoT to ensure customers achieve actionable insights, optimal usability and improved performance and security.
Three Reasons to Head to the Edge
When designing an IoT solution, enterprises should keep in mind the main ways edge can be leveraged to improve performance. Werner Vogels, Amazon’s CTO, famously outlined his three “laws” or arguments for local data processing: the law of physics, the law of economics and the law of the land.
The first argument is to do with speed. Physics dictates that the time taken to send data to the cloud and back will incur latency, but local processing can go a long way towards eliminating or at least reducing that time-lag.
When it comes to data processing in the cloud, time is money. The second law highlights that bandwidth expenses can quickly rack up if you are sending too much unnecessary data to the cloud. Better, then, to use locally-placed intelligence to filter only the high-value data for storage and analysis in the cloud.
Thirdly, the law of the land reminds us of the regulatory and compliance obligations associated with data gathering. It is often necessary to process data locally for legal reasons.
Infostretch agrees wholeheartedly with these three reasons to adopt edge computing, but we also suggest that enterprises need to consider three further factors too.
1. Customer experience as competitive differentiator
Customer experience is an important driver for edge adoption and customers want seamless experiences that just work.
The onus is on enterprises to ensure that edge solutions integrate with hardware, other business applications and the wider infrastructure environment. Rather than thinking of edge as a bolted-on solution, companies should instead regard edge as part of the DevOps cycle of iterative improvement.
Testing methodologies will also need to reflect what’s going on at the edge. Techniques such as end-to-end testing and chaos engineering are particularly suited to managing performance in distributed systems.
2. Analytics at the edge
The ability to filter and analyze incoming data from sensors and other devices is one of the big advantages of edge computing.
In-built AI and analytics capabilities can quickly extract value from the vast amounts of data collected. By converting inputs into actionable intelligence, they not only speed up performance, they also ensure that only the most important data is retained for storage or further analytics in the cloud.
Infostretch, for instance, recently worked with a bioelectronic medical technology company with a pioneering solution transforming the treatment of essential tremor.
For a long time, the only way to treat this chronic condition was either a partially-effective prescription or brain surgery. Using the company’s electrotherapy-based nerve stimulating wristband with comprehensive monitoring and tailored diagnostics, patients can receive treatment that’s calibrated to treat their specific tremor symptoms – for context, the average duration of a short-term tremor is 96 minutes, 75% of patients reportedly experienced meaningful symptom improvements after a single 40-minute session.
3. Hybrid environments
The reality is that multi-cloud adoption is rising and the use of hybrid environments is common, so for many enterprises any edge solution they consider needs to be able to perform across multiple systems. Several approaches can work successfully in the hybrid cloud model so long as careful consideration is given to how data partitioning and processing are designed.
Leveraging Edge for Improved Business Outcomes
Edge computing adoption is on the rise, driven by the multiplier effect of increasing IoT and M2M communications, 5G rollout and digitization triggered by the pandemic. The business case for edge computing is solid – edge solutions increase the speed of digital services, reduce latency and save money, while helping ensure data is processed in full compliance with local regulations.
Enterprises are also increasingly looking to the edge as a source for innovation and competitive differentiation. To succeed, they need to keep focused on the business results they want. Selecting solutions in the crowded edge marketplace can be challenging, but focusing on business value and customer experience should be the priority.
Ultimately, whether you are an emerging innovator or a leading enterprise, edge computing can help you leverage opportunities and improve decision-making. And that is a win-win situation for everybody.
To learn more about how Infostretch can help solve your biggest technical challenges, get in touch using the form below.