Automating the Edge with Robotics

Edge computing research roundup: The upside is big for MEC – if you manage common challenges

A pair of reports published recently describe the two peaks that edge computing must conquer: convincing chief executives to get off the dime and preparing CEOs for early edge-project problems.

Neither climbs are necessarily inspiring, but edge marketers and in-house advocates can take comfort in knowing that every new chapter of information technology begins the same way.

A report by market research firm Mobile Experts Inc. identifies the rationale for use of multi-access edge computing (MEC) in three industries: furniture making, fast food, and oil and gas.

For example, energy production should be a natural proving ground for edge systems, according to the report. Contrary to popular perception, finding and developing an oil or gas field is not a gut decision. It involves long-term, remote operations involving a great deal of data. Nor is operating a pump a one-and-done routine. Infrastructure is dispersed, sometimes globally, it can demand real-time monitoring over time, and data volumes are extreme. One well creates 10 terabytes of data every day. The report outlines how edge computing can reduce costs and improve operational efficiency in the industry.

There is a catch for organizations in all three industries, however. Mobile Experts says that “Each business will need to develop a personalized edge computing business model.” That means few standard blueprints for success, but spending time upfront to analyze requirements can ensure edge computing projects can realize cost savings and new revenues.

All of which leads to the second peak, illustrated in a sobering Beecham Research report titled “Why IoT Projects Fail.”

The document examines IoT projects at various stages, from pilot (50% of cases reported) to stage one early deployment (35%) to major deployment (10%).

Fifty-eight percent of respondents to a Beecham survey said their IoT project was mostly or entirely unsuccessful and the report’s authors say the real total could be 70%. A mere 12% said their project was fully successful.

With numbers like that, the competitive advantage of a successful IoT development will have to be clear before a CEO will approve and stick with what likely will be a bumpy development cycle.

Survey respondents who ran into trouble said business outcomes, including a definition for success, were not sufficiently thought out. They also tripped over organizational problems like not having the expertise to succeed and not maintaining schedules.

A somewhat surprising hurdle among early adopters was being blindsided by technological problems. The biggest challenge was security. The report states that a “one-off” mindset proved unsuccessful. IoT functions and components continue to evolve, and security needs to evolve flexibly with them.

In other related developments:

Cloud research and analysis firm Futuriom published a report on 5G edge evolution. Key applications for the new infrastructure will include Industry Internet of Things, video surveillance, augmented reality and smart environments. Top public companies to watch include AT&T, Ciena, Dell, Nvidia and Nokia.

Research and Markets has pushed out a report on how facial recognition is using edge computing. Globally, the market for face scanning on edge systems is expected to reach $2.3 billion by 2025, up from $765.7 million last year, at a compounded annual growth rate of 20 percent.

Last, Intel and Udacity are partnering to train IoT developers for the edge AI. The pair want to train 1 million developers. Deploying edge AI could result in personalization features people want from an application without putting that personal information on a server in the cloud. That is a proposition that will fuel growth in this area, according to the companies.

Article Topics

 |   |   |   |   | 


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Featured Edge Computing Company

Edge Ecosystem Videos

Automating the Edge


Deploying AI Models at the Edge


Latest News