good morning everybody my name is Josh Wong I'm the general manager for good orchestration at General Electric and I'm thrilled to be here today to talk about Innovation at the GE news desk in distributec 2023 in San Diego sunny California and I'm just as thrilled here to be joined by Colin today to talk about how we have been innovating the energy sector as General Electric so Colin why don't you introduce yourself thank you very much Josh again good morning good afternoon good evening to all I'm Colin Paris I'm the senior vice president and chief technology officer for G digital welcome to distributec so Colin it's an exciting time to be here now with with everything that's going on around energy crisis energy transition Innovation is really the key to unlock value in this changing ever-changing sector so you have been focused on Innovation right from the get-go at GE and you are really looking at past present and of course the future so where are you investing your time these days and why and what are three things that interest you the most as you think about the term grid orchestration and where it's heading Josh I love the way you frame that around Innovation you know GE has had that in its Heritage even going back to the days of Thomas Edison so for us this is part of of who we are and especially at a transition so let's talk a bit about this notion of a transition right and where we're heading what we tend to focus on on GE are really all about the problems with solving what are the inevitable problems in this transition so let's talk about a few and by inevitable I mean regardless of what you think you have to work on these things are going to happen first thing we know for a fact that we have seen many extreme weather events right we've just had the experiences over the last couple of months right whether it be these prolonged ice storms or we have these atmospheric rivers or we see what's going on with the weather patterns over the last couple of months they're happening so how do we look at that in terms of the energy transition the other thing we're seeing is that we are bringing in more renewable sources and these renewable sources are variable whether you bring them on a centralized side or the distributed side right so we've got to deal with that inertia and third we're looking at the fact that many more people are moving to this one sense is that there's a sense of altruism we need to help the planet the other is that I want some measure of energy security I would like to have the ability to generate my own energy when I need to so those are the three problems whether you look at weather inertia or the deals expansion that are happening and we're looking at them because they're inevitable the other thing that's inevitable is data so so here's where um I think data has a unique part to play everything I described to you right now has a high degree of variability the weather is variable right when we add a lot of deers we're adding them they're distributed in a way that we don't fully understand that's variable even with centralized intermittent generation it's variable variability is one thing we're looking at that disturbs the flow that we have in a normally deterministic way volume we're seeing many more deals being added velocity we're seeing the rate of change by which all these storms are happening so when I look at the power system we have and then the variability the volume and the velocity I've got to find a way to factor all of that in in a way that makes sense for the future and that way to me is to data an AI and a useless domain or the domain knowledge and models so this is the motivation for us it's a perfect sort of place to innovate you mentioned to me three things well those are the exact three things I want to focus on I want to focus on how do I actually use the data I have in weather situations maybe I could actually begin to think about is there ways to do outage planning and outage prediction using data and AI to help me with the weather events in terms of inertia is there where ways I can use the models that I have and actually use these AI systems to find ways to occasionally balance you know where I can manage in Usher rather than on a global scale and then in terms of the deals how do I get the right way to represent those loads with their variability so I think this is a time when Innovation is an imperative as well we in gec that is as its Heritage in venovo we have many places to go for sources of information that support us as digital you know so I think it's a great time for us to be here now with that same theme in mind let me let me ask you because you have a Heritage with Opus One of innovation yourself this is your brand so so tell me a bit about this notion of orchestration happening it's happening everywhere we talk we've got orchestrate and pull these things together tell me your definition of that and how do you think of that in terms of where we go in the future perfect thank you so much and I think you hit the nail on the head which is around data and analytics and with increasing unpredictability complexity and volatility in the system so we talk about DRS as one of the major factors right we want the to drive the energy transition we need to flip the pyramid upside down so instead of top down generation transmission distribution we have to decarbonize by electrification and we have to decarbonize based on decarbonizing the edge and empowering The Edge to participate on the grid as really that platform that backbone of the Energy System but now that completely changes the Paradigm of grid operations right in the past very predictable one-way top-down power flow now we have many many millions of sources some controllable some uncontrollable some predictable some unpredictable some can be a top-down type of hard scada command and control signal some you're going to beg it here's a Starbucks gift gift card to incentivize a solution some you can directly interact as well and some you got to go through aggregators and now let's remember the the most of us most prominently increasing Trend which is economics is going to drive how ders would operate in the near term I would say so the whole notion of dsos or transactive energy and having economic pricing signals that relates and Associates with the value that DRS can provide on the grid or not provide in terms of a negative pricing this is key to really managing the complexity in the system now that cannot be traditional management of the grid so I think we need to have a fundamental word change from managing the grid to orchestrating the grid and what that means in my opinion is really going from top down control to System of systems intelligent coordination it involves Utility Systems it involves customer systems it involves behind the meter involves a virtual power plants wholesale markets so all these systems need to work together and utilities may or may not have direct control over them but and they would actually dictate and govern this stability inertia operationally the congestion of the grid even more than traditional centralized generation so how do we wrap our head around that right so very soon utilities will be an orchestrator either a broker or Central coordinator they still need to maintain the fundamental Mandate of maintaining reliability resiliency on the grid they need to serve as customers and customer Choice as they partner together in this decarbonization process and ultimately because they are regulated they still need to remember that they have to keep affordability and cost potency of the system so that type of multi-party multi-system that is the future that is inevitable it's a mega Trend it has to happen and that's why we have to enable utilities as GE to actually have the solutions to comprehend a wrap around all the complexities in the system so I think the the magnitude of disruptiveness is one the other is really I think our job which is I think the the temp the time that we have is running out so it's really time for us and and us and the Innovation team to accelerate the not just tinkering with um with low TRL or early stage technology projects but how can we accelerate the the productization the commercialization the availability and I think that's perfectly going to be our role perfect I love the way you frame this especially this word orchestration right so in my head there are two things and I want to get you a view on it one is all about flexibility flexibility to me is you know I have the capability to do something but things are varying volume and other things right so you know when an orchestrated grid it could be flexible when conditions change the other is adaptability when something new happens I've never seen before how do I put that in and it seems in your term you've gathered both both the flexibility as well as the adaptability am I right it has to be it has to be because that's that's these are the realities or the forces that are either we call a threat or opportunity in this system right so I think and one more thing I would add is actually scalability too uh largely because if it's small amounts of flexibility I think traditional means yeah you can brute force it adaptability we can keep running the same power flow over and over again but now innovation has often been in this on the show floor in the past 20 years that we have been in this space really classified as tinkering projects right it's a pilot here a pilot there we always say we pilot to death but now is we we are running a show out of time and we are running out of uh out of pilot type of limitations and and very soon we have to go from the the early mentality to now we have to scale this as the norm but I think this is the year that we want to transition out and say make a steak in the ground and make a claim that says today this is the year that we have to scale yeah perfect perfect you also mentioned this notion at time too which is always interesting to me right so we um in innovation of GE we have multiple time scales yeah things we're doing with products now there's ways we're looking over the next two years doing pilots and then we have research programs that have been a part of you've been a part of in which we're getting investments from governments to do things and this notion of time allows us to deal with what's coming ahead in the future and to respond you know give me an idea you know from your perspective about some of these time factors that you're seeing are there things that you're seeing on the journey that are coming sooner uh things we should look at things further on so so the the urgency to change is now where are we going to focus on the change is highly relative to these particular utility and grid right it's dictated by demographics customer mix generation mix regulations policy Etc so I think that goes back to the agility versus agility and adaptability versus the scalability side we scale too fast and then I call it throw away code right if we scale too slow and keep agile then we continue to be in Pilot land for too long so we need to and I think that's why I'm excited that the whole grid OS announcement which is now here's a a flexible scalable platform that we you know we know can orchestrate data sets from various sources from various parties in a common data Fabric in a unified Network model of the grid the digital twin of the grid and in a zero trust security trustworthy type of sense and now the applications the the use cases that we can overlay on top has the ability to be agile enough to walk with the utility despite what where they have to focus on in near term but now gecs ourselves as the partner in in addressing some of these early use cases but also a partner along the journey from one use case to the other and to the third and the fourth Etc and it's a notionological progression not siled isolated projects but a journey towards the energy transition and we are ready with the staying power with the resources to invest and with the the ability to scale so that we can walk a very long journey with our customers perfect perfect okay great great great you know one other point I mean this word orchestration also you know has me thinking about an actual Orchestra in which the silos you know of different types of musical instruments but they come together to give you that Harmony and something that's built out of that right you mentioned multiple data sources so there's new data coming in that we're going to be pulling in right both real time and non-real time you know in your opinion as you look at some of those things what types of data sources or is there a way we should think about this data that's coming in you know in order to use it more effectively yeah so so orchestration is is a great example or the the orchestra and I used to be jealous so I love how how we are we are so it's how we bring all these sources together and working in harmony but the orchestration first started I would say within a Durham space right that's where I came from because we have the solar and the storage and electric vehicles and smart thermal stats and a building management systems I think there is no end in slight to the diversity and variability of the asset mix that's happening at the edge right now but what we realized I think very quickly is derms provide the orchestration of the Dr portfolio but we need to go beyond that to look at the grid because we need to make sure the grid is safe secure year affordable reliable resilience Etc so now we need to start introducing other data sets on top of the Der data sets one is weather the other is demographics we need to bring in of course the markets the markets are going to dictate and now wholesale and Retail distribution of transactive level type of markets I think those are critical and I would say one of the biggest greatest thing beauty of innovation is the law of unintended uses so we're both laughing because we know how how how how opportunistic that can be so when we expose the data set when we can mine design patterns such as mashups of various data that you and I have never even dreamed of and we opened up that ecosystem to third parties and partners and utilities I think there's no limit in sight and I'm very very excited to see what new data sets people can overlay on this grid yeah yeah I think you're right too I mean and the other brilliant thing is that we're part of GE van over so we bring in domain knowledge and people who do generation from other aspects you know both about the variable as well as the you know the gas power type of generation and then we also have some great energy consultant and other people so we're bringing all these pieces together in a way that fits in this notion of orchestration totally and I think that that's why we call it the orchestration a good orchestration but ultimately the grid is the backbone or the platform for the energy system in total and so let's highlight like one or two use cases an example right electric vehicles I think that's a Hot Hot Topic today and but electric vehicles is not just a derms problem or an adms problem with enough penetration of EVS you have Dr portfolio aggregation forecasting which becomes very challenging with EVS given there's a huge behavioral and traffic pattern component that's another data source that we need to start bringing in then we have the grid orchestration side on all the congestion and switching outage situations the risk factors with hosting capacity but now EVS will go all the way up to wholesale and up to the bulk power system are our for example our nuclear Fleet going to be flexible enough to match the flexibility of demand like do we need fundamental innovation in our base generation our our Renewables can we we cannot dispatch for example solar win but we can tweak the power factor we can tweak the way that they they inherit they build um inertia in the grid so I think all those pieces have to work together another example is like storm preparedness outages right I think part of the study that that this team our team has done is to show that if we bring tnd coordination and der coordination all together we can restore 25 to 40 more customers in an outage and I would add to that right the restoration capacity increases because of ders the ability to know the line ratings and get much closer to capacity is also much higher as well so I think that's really where we need to think outside the derms box or the adms box but look at here's the grid here's how our utility customers and partners oversee the grid how do we help them orchestrate all these Data Systems together to protect the grid I think it makes a ton of sense to me I mean I love the last example you had about the tnd coordination in terms of you know outage risk understanding and planning and restoration the other thing I love about that is that it's building upon the new capabilities that we see showing up with data machine learning and AI it was interesting to me that a lot of the work we were doing on forecasting you know we're using a lot of VI models to look at the historical data and the impacts of what happens you know during the storms you know what lines get you know affected the most yeah we're also using the same thing to understand the loads that are going to be happening so there's AI both in the forecasting of generation forecasting of loads the prediction of Storms and then these AI used in optimization right now that can get it done faster more accurate and in some cases you know even at lower cost so I think there's a beautiful combination of the data AI capabilities with the problems we're facing in orchestration that will have a unique output for us totally agree 100 because 150 right um and I think uh I think us coming from the engineering the classical engineering space we are going to see or we're already seeing some of the computational limitations or just pure algorithmic limitations um of physics or mathematical based optimization for example to be able to catch up to the to the agility and and flexibility of the grid for us to go to the next disruptive step function we need Ai and I think looking at back in the in the startup space we see you know I saw a lot of AI companies have great ideas but they cannot really tweak the use cases of this grid because running a degree is complicated but now I think we we are on the verge of a fundamental inflection of AI in this space and so my last question to you then is rating from 0 to 10 how far are we on the AI Journey as an industry within the energy sector you know I will tell you it all depends upon which price process you're looking at if you're looking at the planning part if you're looking at the operations if you know maintenance there's so many different processes we look at right but in all of them I would tell you I think we are on the three to four space you know because we we are very early to this we have to get the data right and then we have to actually get the fit you have to find the Fitness in in terms of is the problem that we're trying to solve fit for the use of AI and domain coming together and then the other thing is that every time you introduce probabilistic things you introduce uncertainty and we've got to educate people in our culture in the energy space on how you look at uncertainty so I think we are well on the way you know but at the early stages but man it's profound and we would gee we're going to be here to yeah walk you through all the way exactly it has been a delight Josh to have this conversation of course as always and thank you so much always great talking to you and always great learning from yourself my pleasure same to you so thank you everyone and you have a conference and more to come
2023-08-26