AI-supported automation can enable the use of carbon capture technologies: Luiz Gonzalez, AboitizPower – ETCIO South East Asia

Luis Gonzalez, Chief Data Officer, AboitizPower & GM Data Innovations Power Industry, Aboitiz Data Innovation.The role of automation is pertinent to newly rising tech, especially in ever-growing and unpredictable industries such as power and energy. Delivering optimal results for the increased industrial demand of the energy sector requires the deployment of next-gen automation powered by artificial intelligence and machine learning. The novel entrance of robotics and automated solutions will help businesses in the power industry to inculcate growth and stabilise efficient consumption.

To learn more about the emerging technologies the power industry is adopting, the sectors challenges in meeting sustainability goals, and the role of automation in furthering its scalability, we interacted with Luis Gonzalez, Chief Data Officer, AboitizPower & GM Data Innovations Power Industry, Aboitiz Data Innovation.

Luis leads the experimentation and development of AI solutions in areas of key importance like power sustainability and positive societal impact. His expertise is backed by his extensive experience in digital transformation and machine learning adoption. He focuses on advisory and transformation on how to extract value from machine learning, ERP, analytics, IIoT, GIS, control systems, and other AI solutions for enterprise clients in Asia Pacific, North/South America, Europe and the Middle East.

Here are the most interesting excerpts from the interview:

In the various emerging subsets and derivatives of next-gen automation, what is the role of automation in the energy sector?The mandate to deliver power reliably, economically with low impact to the environment is an increasingly complex one that requires the orchestration of multiple partsfrom fuel futures, power and grid operations to consumption anticipation.

Artificial intelligence (AI) is a powerful tool that is used to aid in the orchestration and forecast demand as well as manage the distribution of resources, to ensure that power is available at the time and place it's needed with minimal waste and impact to the environment. This is especially important in the renewable energy industry, where energy cannot be stored easily and optimising the power demand, time and location of where it is generated is particularly challenging. AI will also play an essential role in the worlds transition to clean energy by providing more accurate forecasting of supply and demand.

Maturing technologies such as cloud, also play a significant role in dismantling barriers to connectivity, scalability, and effective data management to provide seamless and instant connectivity and computing power that is scalable and at a lower cost.

Luis, your vast experience in deploying AI and automation in the power industry is commendable. Kindly elaborate your observations of the use cases of automation and the evolution of its developing technologies across the years in the power industry.For decades, we have implemented sensors and statistical approaches in control systems within the power industry. The real challenge, however, has been the inclusion of contextual responses and the interpretation of thousands of alarms. Most modern power plants are oversensored and lack real understanding of signal reach, which is a key aspect that maps decision making. The power equipment manufacturers, such as original equipment manufacturers (OEM) like Siemens or GE, have invested in the data collection and life prediction algorithms for the equipment they manufacture. However, the progression of AI in power requires that we build machine-led inferences and human-assisted monitoring and optimisation in a Transfer Learning approach where the AI models become more accurate as the human experts use them. It is early to say that full automation is possible, given that there is not enough understanding of the disruptors yet in our industry. The disciplines to make this possible already exist, yet only the academics who explore deep learning understand them well enough.

Currently, there is still a gap between the number of people who favour the acceleration of AI adoption and those who are operationalising solutions, over 80% of the region still in the early stage of AI adoption. My perspective is that, as the region accelerates its digital maturity and more AI engineering capabilities grow in startups, advisors, software solutions, operators and executives, we will begin to see a roadmap for true automation of power plants and grid operations in an orchestrated manner. As every grid is different and regulations governing the industry differ by country, it is very likely that each country will need to localise and develop solutions that work for their own market.

The energy sector, while undergoing rapid transformation, is navigating ways to combat the growing challenges of sustainability. How can automation ensure organisations meet their sustainability goals?Let's start by considering the current system of centralised power and multi parties grid operations, which produces inefficiencies in the distribution of megawatt hour (MWH) to where they are needed in time and demand. Furthermore, centralised power has significant benefits in generating at a lower cost, but the secret to begin reaching sustainability goals is conserving and delivering all generated power as best we can without losses. Schemes of power storage, efficient distribution optimisation, demand management, and distributed generation are all facilitated by data to define the where, when and how to best run said schemes with minimal CO2.

AI-supported automation can help organisations meet sustainability goals by providing ways to reduce emissions in power plants, enable carbon capture technologies, reduce outages and facilitate hybrid systems. Not only can AI predict potential volatility within the entire system, but it can also better identify efficiency gains through optimisation of generation, management of grids, placement of storage or sharing of power between generators or consumers.

We have seen how automation can help reduce carbon footprint through machine learning and predictive models. One of our successful use cases was the 28-Day Predictive Strength Model developed by ADI and Republic Cement. AI was used to predict cement quality, effectively optimising concentrations of raw materials, saving time and improving product quality through consistency of cement products. On the other hand, ADI and AboitizPowers machine learning model reduced the consumption of limestone and paved the way to explore new areas for operational excellence in one of its power plants.

How are renewable energy companies in APAC deploying intelligent automation and robotics? Give us a brief overview of what lies on the road ahead for automation in the energy sector.In the renewable space, weather prediction and early impact modelling is key not only to optimise potential power generation, but to also minimise expensive maintenance and maximise investment return. Robotics is not necessarily as present yet, however, we do have drone technologies aiding in the maintenance and inspection process of aeolic and photovoltaic fields. The real benefit for renewable companies from the adoption of learning algorithms, will be the facilitation of hybrid systems that can work with fuel based and renewable based generation as well as store power for grid resilience, peak load management and reduced CO2 and cost of operations.

Contrary to popular belief, the reality of our aggressive sustainability targets are not going to be achieved by finding a new energy intensive and zero carbon emissions fuel that can be adopted everywhere. This is wishful thinking as we are very dependent on our infrastructure to be able to sustain our demand for power and economic development. Rather, the way to achieve our goals is by designing and adopting hybrid systems across our grids and leveraging data to better understand, predict, simulate and prescribe these hybrid solutions, so as to incorporate the existing investments we have done in infrastructure, while replacing faster the more wasteful and inefficient parts of our grid.

View post:

AI-supported automation can enable the use of carbon capture technologies: Luiz Gonzalez, AboitizPower - ETCIO South East Asia