Intel demos Energy Collective Optimisation [ECO] Edge Demand Response system in Las Vegas at AWS re:

We often read about IoT for the industry and the smart home, but hardly ever see the combination of both coming together to provide a very large scale autonomous solution. Let me give you a bit of background how (energy) Demand Response, IoT, EDGE computing and a bit of industrial design work towards deploying an active grid management system at the edge

Homes, appliances, EVs, machines, industrial equipment, all of them are becoming smarter, connected and richer on features, yet there’s one missing link: collective energy management.

Why a Collective Energy Management system is required? And why should you care? Let me break this into 4 parts:

The Load, the energy use at the edge (customers, factories and people’s homes)

Electric Vehicles (EVs)

To put it simply, the electrical grid cannot take infinite amounts of energy loads. Think about the ‘cul-the-sac’ street in your neighbourhood, how many electric cars are there [EV]? Probably not many if any at all! I’m sure you heard about peak demand times for the grid, and how usually around dinner time a massive peak of energy demand appears on the grid, bigger than what the factories and industries use together. When you take into consideration the millions of homes, even at small loads like 2-3kW, aggregated together, it adds quite a bit.

Now replace the 2-3kW of energy for EVs in a few years. A EV will use +3kW on a typical home socket, however it will take several hours to charge as EV batteries are becoming normal to be around the 30-50kWh. That translates in 10 hours charging for a 30kWh battery. Now imagine that high demand when people prepare dinner at home, the usual 3kW that puts a bit of stress on the electrical grid and multiply it by 10. Hopefully by know you get where I’m going with this…

EVs can also have faster and rapid chargers, meaning that are able to charge a 30kWh battery within a couple of hours. That means even more stress to the grid, why? We humans follow very similar patterns, the typical 9am-5pm work, arrive home and plug the car so it is ready for next morning. Add the usual already existing demands for every days life to this.

HVAC – Heating, Ventilation & Air Conditioning

It is well known that so far the highest need for energy on domestic use is for HVAC until the EV arrives at least. When we talk about heating we often think about gas, however a big part of heating comes from electricity, this is even more common on high rise flats where the logistics and safety on gas are supersede by electric heating. Then we look at cooling, this is an electrical load too. As the world population keeps increasing, so the requirements for HVAC.

The (electrical) Grid

The obvious solution is to make the grid bigger, put bigger generators or even set new localised generators closer to the point of consumption. All of these are good solutions, just not fast enough and definitely not that easy to implement due to a high level of disruption, cost and most importantly waste. Yes, waste.

The grid has fluctuation on demand, that is: the grid maybe overloaded at 6pm when everyone gets home, plugs their EV into the grid and the heating or air conditioning is switched on. However there’re other hours during the day where there’s spare capacity where there’s excess of available energy.

Wouldn’t be great if we were able to flat out the load? Take the high load peaks that push the limits of the grid and move them into those time slots where there’s spare capacity? That is Demand Response, the simplified image below illustrates the power of demand response:

As you can imagine, managing the electric demand of a few thousand system is complicated enough, let alone millions of homes.

Energy Collective Optimisation engine for Demand Response at the Edge with Machinon

This is where everything comes together, where Cloud meets the Edge, or better read where AWS (Amazon Cloud) meet computing power at the Edge (Intel).

Controlling loads at the edge, without the cloud? Well you need the cloud to synchronise and aggregate the energy the same way an Orchestra needs a Director, however that does not mean that if the Director disappears for a few seconds, the Orchestra wouldn’t be able to continue playing.  This is very important, to understand that the cloud add value and the edge needs to be self sufficient to take decisions locally based on the last guidance from the cloud plus of course some AI at the edge. Merge two behemoths’ technology like Amazon (cloud) and Intel (edge) at massive scale and you have the solution.

So what does a small company like Logic Energy has to do with Machinon with these to behemoths of technology?

Simplicity and Open Architecture. Logic Energy has been working nearly 2 years developing Machinon, an open Smart Home and Automation device that merges both: the domestic use with the energy (industrial) needs. Let me go into a bit more detail what I mean with it with some details:

Computers are really powerful and fast, however they tend to be a bit hopeless when interfacing with the real analogue world. Not saying they cannot do it, it just is a lot of hard work to do it and besides not the best use of their skills.

On the other hand, more embedded and industrial semiconductors are much better to interface with the analogue word, some of them are real tanks when comes into surviving harsh environments and interferences that a computer would not be able to survive.

Now join both, the powerful EDGE technology from Intel with a tough and hard designed embedded design for the analogue (real) world and you have Machinon. Below is an image of the demo kit used at Las Vegas re:invent 2018 on Intel’s booth, you will see that Machinon resembles a lot an industrial PLC which usually you don’t find inside people’s home.

Wolf in sheep’s clothing

Let me take you back to the grid, remember I was mentioning that upgrades and changes to the grid are not easy and can take a very long time (many years)? If you merge the grid with domestic Demand Response, you want a solution to last many many years without any kind of technical support. Unfortunately wireless networks are not like that, don’t believe me? Compare the men workforce of any mobile / cellular network vs the workforce for maintaining the grid. Yes, I’m talking wireless vs wired, but don’t get me wrong, I like wireless too as it is very convenient and this is where we merge “domestic” technology with “industrial” technology: Machinon with its open architecture gives you both options.

If you want something to stand the pass of time, which usually is the case for electrical grids and anything related to it, then probably you are better of using wires.

If you want your customers (domestic users) to add smart home appliances, thermostats, add rule engines for their own requirements… you can too via wireless.

This is my point about combining multiple industry areas and keeping an open architecture as every customer (you, the energy utility company, the grid operator) has different requirements and all of them need to live under a single solution.

I tried to explain what we are doing together with Intel right now. Our role right now is to support Intel with what we call “back to back” (cloud to edge) working solutions. We are providing a consultancy support to Intel for this, and with one key different to the usual consultant: we are actually delivering the product too! We have been 10 years delivering back to back products.

As a final note I would say that the world of Demand Response is big and complicated, and when we move into Domestic Demand Response… it becomes a lot more fuzzier and complex. I’m more than happy to discuss things offline if you are interested, feel free to contact me through my LinkedIn profile.