For the industrial sector, AI is a set of tools and techniques that can be used to drive lots of optimisations across different disciplines and different markets. Generative AI is being used with design tools to let designers of industrial plants, design things in a safe and sustainable way. By looking at patterns of previous designs it can make clever assumptions about how to optimise future designs, explains Rob McGreevy at AVEVA.
From a macro perspective, energy markets have two challenges. The demand for energy is increasing dramatically. Just in data centres alone, it is like three times. On one hand, we have this incredible increase in demand for energy, while a significant part of the population has no access to power.
At the same time is the need to have more sustainable energy, which is decarbonisation and alternative energy sources. So how do we meet the energy demand and at the same time, decarbonise and diversify the energy sources that we have.
The big opportunities for enterprises are through what AVEVA calls its handprint, which is the ability to help customers reduce their carbon footprint, and help them become more sustainable, which is a focus for AVEVA. There are an enormous amount of use cases around for using less water, air, steam, gas, electricity, and decarbonising.
The first thing most enterprises want to do is capture data around water, electricity and gas flow rates, and then from there, you can begin to drive optimisations. And this is where you start to apply AI analytics. What are the reasons we are consuming so much water on this sort of steam turbine? And then the analytics can point to ways to reduce the consumption.
These are well documented cases and AVEVA has got cases where turbines can save 100,000 gallons of water purely by running an optimisation. AVEVA can reduce CO2 emissions by optimising the fuel that is burnt.
What is industrial AI?
AI is a headline topic and is probably the most transformative thing that will affect industries in the next decade. Specifically for the industrial sector it is a set of tools and techniques can be used to drive lots of different optimisations across different disciplines and different markets.
AVEVA is currently using Generative AI on authoring and design tools to let designers for power plants, offshore oil rigs, whatever it may be, design things in a safe and sustainable way. And what Generative AI does, it looks at the patterns of all previous designs and makes clever assumptions about how to optimise future designs.
If I am a designer and I am trying to connect four tanks with four pumps, rather than me drawing the diagrams out, drawing the piping, the tool itself will make a recommendation on the right way to have it and they can make decisions on the cheapest routing using different pipes and materials, and the most efficient design.
On the authoring and design side, AI and Generative AI is used to help designers create sustainable, safe, optimised engineering initiatives.
If you think about AI and the production operations side, lots of amazing things have been happening there. That is a more mature market. AVEVA has been doing AI there for almost two decades. This is the realm of predictive analytics and prescriptive analytics.
AVEVA looks at equipment like turbines or compressors, measures all the variables and looks for anomalies and patterns in that. These different behavioural changes in the equipment are sent to operators and maintenance technicians as notifications, that there might be a problem with this equipment. That is predictive analytics.
Prescriptive analytics becomes not just notifying someone that there may be a problem but advising them on what might be causing the problem so that they can actually take corrective actions.
Roadmap for implementing industrial AI
AVEVA creates templates for different types of equipment that exists in manufacturing, but all plants and all operating environments are unique. This is where customers themselves, or a system integrator will help the customer tweak the models a bit based on their operating conditions and their equipment.
This is really a tight partnership between AVEVA as the technology provider and the customer. The data is always the customer’s data, so they use AVEVA tools to capture high fidelity process and production data. They use AVEVA models to be the starting point for analytics. And there is a lot of great cases of this.
Customers would start with using AVEVA process historians, and then look to drive improvements. And then the vendor would come in and sell them a predictive analytics capability. This is where AI that sits on the top and helps customers identify anomalies, optimise for fuel consumption, or fuel burn, among others.
AVEVA supports both on-premises and in cloud and call this a hybrid solution, so customers can have some of the historians on site. They can choose to connect into the CONNECT platform in the cloud, and they can do both, if they like. If they want to just do the cloud, they can do that as well.
For most process heavy industries, especially ones that have safety and reliability requirements, the hybrid model is the most popular. That is some stuff on-premises and then using the CONNECT cloud platform to do long term planning, analytics and visualisation.
Today AI is in some but not all AVEVA products. Eventually it is going to make its way into all of the vendor’s products. The most notable one is the industrial AI assistant, which provides operators and users of the system guidance on how to find information, access information, and is almost like having a conversation.
The data is sourced from any customer automation system, control system, engineering design documents, alarms, events, process anomalies. You are having a conversation with an industrial assistant.
Role of channel partners
The basic industrial AI assistant works with AVEVA tools and there is not a lot of custom work that needs to be done. On the predictive and prescriptive analytics there are sophisticated models. This is where system integrators or specialists from the AVEVA team, do a bit of work, because there is more science and math involved for tweaking the models.
If customers are looking for a reliability centre, maintenance, predictive analytics, prescriptive analytics, anomaly detection, AVEVA solutions can do a lot of that. It starts to get sophisticated if you are doing advanced thermal modelling for example.
If customers are on the AVEVA CONNECT platform, they have access to the tools and technologies. The platform is AVEVA CONNECT, and the commercial model is Flex where they can buy credits. And if customers have Flex credits, they can use those against any of the technologies in the AVEVA portfolio.
