Skip to main content

When I hear the word “smart”, my first reaction is the traditional definitions of “sharp”, “intelligent” or “showing quick intelligence”, but in the “non-human” context, it is defined as “the ability to function with minimal human intervention or independent of external control” or also “equipped with, using or containing electronic control devices”.  Following the latest non-human definitions, you hear about smart things everywhere, about how digitalization is changing everything, about the era of interconnectivity….

In the electricity sector, it’s the same trend: “smart cities”, “smart meters”, “smart grids”, “smart chargers”… and everybody is talking about digitalization and interconnectivity being the key to being “smart”.

 

My problem with the word “smart”, I think, comes from my mental reference to the human context, and its relation to “intelligence”. In the literature, depending on the authors you can find between 4 and 12 (maybe more) types of intelligence (logical, musical, intrapersonal…). Therefore, referring to something as “intelligent” is confusing to me, maybe due to my limitations, as defining intelligence is already quite complex and covers a wide range of topics.

 

Many things that today are called “smart” under the above definition are simply sensor(s) that send signal(s) and the signal(s) are translated into an action. Some years ago, this was simply named “automation”. In fact, “automatic” is still a synonym for “smart”, but less fashionable.

 

The only difference between the old automation and in most of the cases, today’s “smart” is that today’s computing power makes a reality to receive more signals than before, and to translate them into more actions in less time…

When talking about intelligent or smart, I always think of how the human brain works.

–  One part of the brain called brainstem functions in automation mode, regulating certain involuntary actions of the body.

–   Other part of the brain, the cerebellum works in self-awareness mode, providing for example control of movement, and balance to our body.

–   And the last part of our brain, the cerebrum formed by the four lobes, providing reasoning, planning and comprehension among other things. This last part of the brain is one of the differences between most animals and the human species.

 

In the electricity sector, companies are investing massively in sensors and in automation that allows them to solve the problems they face as a way to bridge the digitalization gap. Many of these solutions would fall in the automation side of the brain. Other solutions that companies are incorporating emulate certain levels of intelligence through the use of bigdata and artificial intelligence techniques providing a kind of self-awareness side of the cerebellum, and even some part of the cerebrum.

 

Still, in order to adapt to a constant changing environment in an efficient and resilient way, the brain of the network requires something else.

 

At Plexigrid, we believe that to manage the issues that the power systems are facing, the grid needs to be smart in the human context; this means to behave as the human brain, not only in the automation context, or in the self-awareness context.

 

The smart network needs to receive signals from sensors and decide: i) if it needs to react directly, in automation mode, ii) If it needs to work in self-awareness mode, keeping the system stable, or iii) if it needs to interpret, think and plan about what is happening.

 

To do this last part, the “brain of the network” has to model all components and pieces of its own network holistically and not in an isolated way as decoupled processes. The “brain” has to be able to generate different possible scenarios to understand the outcomes of the potential actions, and then take the right action at the right time.

 

This is a slightly different philosophy to the traditional way power systems have operated until now. The “automation philosophy” based on the principle of action <-> reaction is valid when the number of variables is reduced. In an era in which we deal with millions of signals and data, the “intelligent philosophy” based on decision making from complex models is the only way forward.

 

In Plexigrid, this is the way we conceive the network as “smart”. Still, we may have to continue using the term “smart”, or maybe we could differentiate ourselves calling it “neural grids”, or maybe, we should just call them “plexigrids”