Last week’s report explored what IoT datasets can and should be run through ML frameworks and algorithms. Once that is answered, we must determine where they will be run in your IT infrastructure.

 
 

Report highlights:

 
 
  • Where is IoT getting deployed: in the central cloud or on the Edge?
  • How is IoT Edge adoption trending?
  • How do these trends in AI/ML and IoT impact M&A prospects?
 
 

Central Cloud or Edge?

 
 

When discussing where advanced analytics might take place just a few years ago, the default answer was ‘in the central cloud’ given the scale of computing resources required and the fact that cloud service providers have readily made tools available that can run on their infrastructure via PaaS or SaaS.

While a centralized IT environment is ideal for big-data analytics and training workloads, the edge (computing resources closer to IoT edge devices) is an increasingly viable alternative, especially for real-time analytics or applications that require immediate action.

Today, we have the ability to run ML algorithms on devices like Raspberry Pis, directly on devices such as vehicles or traffic cameras, or at some other location with more local proximity to the cloud.

Google has announced that it has miniaturized TPU chipsets to the extent that four can sit on the face of a US penny, enabling almost any device to run ML workloads locally.

ML will run ‘everywhere,’ but more importantly, these workloads will run where it is most economical and technically feasible to support the task at hand.

 
 

How is IoT Edge Adoption Trending?

 
 

Our survey reveals that AI/ML analytics done at the IoT edge jump 5% when comparing 2017 versus 2018.

 
 
Types of IoT Analytics Done at the IoT Edge
 
 

M&A Predictions

 
 

The strategies and tools to democratize the power of AI and IoT form the basis for a hypercompetitive and dynamic supplier ecosystem that includes cloud providers, OT players, IT players and specialists alike.

Therefore, it’s no surprise that AI and IoT are the number one and number two themes for M&A activity over the next 12 months. We expect record deal-making in 2019 given the value at stake.

 
 
M&A Outlook: AI and IoT
 
 

Conclusions

 
 

AI, in all its forms, and IoT are two of the most important technology enablers impacting the human race. When they converge, real magic happens, because AI and ML excel at processing data where humans simply cannot, and the sheer size and complexity of IoT data dictate that, without advanced analytics, IoT data is probably not worth the effort.

Enterprises are just now beginning to recognize the transformation possibilities of applying ML to the data that runs through and around their business, including, of course, IoT.

Both AI and IoT share underlying enablers such as the powerful computing capacity available in cloud and edge platforms.

We have entered a wonderful era of democratization of both, via easy-to-use platforms that allow non-data-scientists to work with AI or business applications that come with ML and IoT data as part of legacy IT/OT applications such as enterprise asset management or factory floor control solutions, or as part of a predictive maintenance service on an industrial machine.

The types of use cases at the intersection of AI and IoT include incredible innovations such as Wi-Fi traffic signal noise to create motion intelligence, worker safety solutions, predictive maintenance for machines and humans, self-driving cars, and building management solutions – the list goes on and grows larger every day.

Stay tuned for next week’s report, which will compare IoT vendor options.