Energy’s Future: A Convergence of Technology, Market Forces, and Sustainability

 

Background

The world of energy is going through a transformative resurgence. It was clear from the discussions and interactions at Energy Risk USA, annual gathering of thought leaders, experts and practitioners in energy trading and supply. It is the central thinktank reflecting  energy industry and innovators, federal and state governments, regulatory boards and researchers.

At the 2025 version of Energy Risk USA in Houston, we had thought-provoking sessions throughout the conference, talking about challenges and opportunities, processes and tools to address them. Energy is undergoing a transformative change in landscape. It was clear there was unanimity the future of energy is seeing a transformative change, discussed the many factors behind it and then, exchanged ideas to address those changes.

Energy is seeing unprecedented demand as the remotest parts of the world gets connected to electricity and high-speed communication – mobile and internet. Making things more complicated, there is uncertainty around supply, volatility in prices, challenges in logistics. Add to that the concern for environment and need to transition to cleaner energy sources, you have a boiling cauldron of complexities in the energy world.

This complex world of energy is a prime domain for digital transformation. From data, analytics to Artificial Intelligence and Machine Learning to automation and agentic AI, the world of energy needs all the help we can provide to address the complexities. Digitalization and AI are seen as the critical tools to transform energy and be ready for the future.

Uncertainty in energy

One of the key discussion areas was the impact of policy and geopolitics on energy. The Russia-Ukraine war, the Middle East instability and other political instability are impacting supply lines and creating volatility in prices. Tariff on solar panels and other items manufactured in China will raise prices on renewable power here in US.

US Government policy to go back to drilling is in contrast with European push for more clean energy. US wants to raise production; OPEC is also supplying more. There is talk that the US government is interested in seeing lower oil prices to handle inflation from tariffs. There is also concern that lower prices will negatively impact domestic producers.

There is short term risk around tariffs and oil prices too, coming from conflicts in the Middle East. Interestingly, Iran Israel was talked about, when it had not started. No one can predict the future with certainty but there was a consensus that there will be uncertainty in supply and volatility in prices.

Future Needs of Energy

To address the challenges of uncertainty of supply, risk, lack of efficiency, complexity of doing business and managing risk, companies must take their game to the next level. Energy companies will be looking for ways to address the challenges:

·         Advanced systems to quantify and manage risk

·         Predictive system based on Machine Learning, Data Science to analyze data and forecast future demand, supply, price and logistics

·         Generative AI systems to understand the regulations, documentation for energy trading and risk management in compliance and governance

·         Profile-based early detection to catch errors in trades

·         Agentic AI to find errors, reconcile and automate front to back-end processing

AI and Digitalization in energy in an era of exploding demand, volatility and drive for clean energy

The Accelerating Demand for Energy

As digitalization expands into every facet of daily life—industrial automation, electrification of transport, and AI-driven infrastructure, the global energy footprint grows exponentially. This rising demand places unprecedented pressure on energy markets, requiring advanced strategies for efficient generation, distribution, and consumption.

Navigating Market Volatility & Price Dynamics

Energy markets are increasingly shaped by a complex interplay of economic forces, geopolitical instability, and supply-chain constraints. The energy transition amplifies volatility as legacy infrastructures adapt to accommodate new technologies. Advanced AI models, capable of high-frequency market analysis and probabilistic forecasting, are crucial in mitigating risk exposure.

Credit Risk & Financial Structures in an Era of Transformation

With heightened volatility comes the demand for stronger risk frameworks—higher margins, collateralization strategies, and real-time credit analysis. AI-powered risk assessment tools enable predictive credit scoring, ensuring market resilience while fostering investment in new energy ventures.

Sustainability & The Reinvention of Energy Systems

The global shift toward sustainability fuels demand for diversified clean energy solutions—wind, solar, hydrogen, biodiesel, and geothermal. In parallel, the optimization of carbon-based fuels like LNG and natural gas remains critical. AI-driven simulations, life-cycle assessments, and predictive analytics help organizations balance economic viability with environmental responsibility.

Energy Storage: The Missing Link in Renewables

Intermittency remains a core challenge in renewable energy adoption. Despite advancements in battery technology, full-scale energy storage solutions are required to stabilize markets and ensure grid reliability. AI-enabled optimization models refine energy storage forecasts, improve resource allocation, and simulate potential failure scenarios.

The Hybrid Energy Model: Integrating Conventional & Renewable Power

A seamless blend of conventional energy sources with renewables is required to sustain a stable electricity grid. AI-powered smart grids leverage deep-learning algorithms to optimize energy distribution in real time, balancing load demand with supply fluctuations.

Investment Strategies for a New Energy Era

Investment in clean energy requires rigorous financial modeling and risk-adjusted decision-making. AI-enhanced portfolio analysis identifies opportunities with high-impact potential, while digital twin technologies simulate long-term outcomes before capital deployment.

The Role of AI, Machine Learning, and Emerging Tech

• AI-Driven Decision Making: As energy supply chains become more complex, AI enhances predictive modeling, dynamic resource allocation, and automated risk mitigation strategies.
• Unstructured Data Insights: Beyond conventional metrics, AI mines publicly available datasets, industry signals, and global events to refine forecasting models.
• Domain Expertise & AI Synergy: The effectiveness of digital tools depends on their alignment with industry expertise. AI excels when paired with deep energy-sector insights.
• Automation & Digitalization: The future of energy is shaped not only by markets and climate considerations but also by the seamless integration of digital transformation. Smart automation will be instrumental in driving efficiency, reducing waste, and creating a more agile energy ecosystem.

Conclusion – energy complexities need digital, data and AI

The global energy sector is evolving into a multidimensional network of interdependencies—balancing supply and demand, integrating renewables with conventional power, navigating market volatility, and addressing sustainability imperatives. This complexity demands more than incremental improvements; it requires a fundamental shift in how energy is managed, optimized, and innovated.

Digital transformation, powered by AI, is the key to enabling real-time decision intelligence, predictive analytics, and automated efficiency at scale. From smart grids that dynamically adjust to consumption patterns to AI-driven forecasting models that anticipate disruptions, these technologies redefine resilience in energy systems. Organizations that embrace this transformation will not only navigate uncertainty but will lead the future—where energy is more adaptive, efficient, and sustainable than ever before.

Leveraging AI in the Volatile World of Energy Trading and Supply

The global energy landscape is undergoing a profound transformation characterized by escalating demand, heightened volatility, and an urgent need for sustainable practices. Artificial Intelligence (AI) and Machine Learning (ML) are essential tools for addressing these complexities, specifically within energy trading and supply. By leveraging advanced analytics, predictive modeling, and Generative AI, organizations can optimize operations, mitigate risks, and drive innovation in a rapidly evolving market.

 

The Evolving Energy Landscape: A Confluence of Challenges

The world is witnessing an unprecedented surge in energy demand, fueled by the accelerating digital revolution and increasing reliance on technology. This surge, coupled with complex geopolitical factors, supply chain disruptions, and the imperative for energy transition, has created a highly volatile and unpredictable market.

 

Unprecedented Energy Demand:

The proliferation of digital devices and technologies has led to a dramatic increase in energy consumption across all demographics. This trend is expected to continue, placing immense pressure on existing energy infrastructure.

 

Unparalleled Market Volatility:

Economic and political instability, coupled with supply constraints and the complexities of the global energy transition, have resulted in unprecedented market volatility. Accurate prediction of energy demand and supply is crucial for ensuring market stability and affordable energy access.

 

Credit Risk and Margin Management:

Increased volatility necessitates higher margin and collateral requirements, posing challenges for credit agencies and regulators. Collaborative efforts are essential to navigate these challenges and facilitate investment in new energy projects.

 

Environmental Sustainability and Clean Energy:

The growing focus on environmental sustainability is driving the adoption of renewable energy sources such as wind, solar, hydrogen, biodiesel, and geothermal. Simultaneously, cleaner fossil fuels like natural gas and LNG are gaining prominence.

 

Energy Storage Constraints:

The intermittent nature of renewable energy sources highlights the critical need for effective energy storage solutions. While battery storage offers some mitigation, significant challenges remain in balancing energy supply and demand.

 

Investment in New Energy Projects:

With soaring energy demand, volatile prices, geopolitical uncertainty and an absence of widespread cheap energy storage options, and variable availability of wind and solar energy, it is important to find new investments in energy. Accurate risk-adjusted rate of return calculations and effective capital management are essential for attracting investments. New models handling future investment, especially in renewables, must be developed.

 

Leveraging Analytics and AI to Address Energy Business Challenges

To navigate the complexities of the modern energy market, organizations must embrace advanced analytics and AI-driven solutions.

 

Modelling and Quantitative Tools:

Sophisticated modeling is essential for accurate risk assessment, margin management, and forecasting of price, demand, and supply. Advanced quantitative tools enable organizations to optimize operations and manage working capital effectively.

Data and Analytics:

Leveraging diverse data sources and advanced analytics is crucial for informed decision-making. Cloud technology and advanced data processing capabilities are essential for extracting valuable insights from complex datasets.

 

Artificial Intelligence (AI), Machine Learning(ML):

AI/ML algorithms can process and analyze vast amounts of data, identify patterns, and generate accurate forecasts. ML techniques enable continuous model refinement, improving the accuracy of predictions over time.

 

Generative AI and Market Research:

Generative AI, such as large language models (LLMs) like ChatGPT, can extract valuable insights from unstructured data, including social media, news articles, and public reports. This capability enhances market research and provides a holistic view of market trends and sentiment. For example, analyzing social media discussions about upcoming events in Asia can provide valuable insights into potential demand surges.

 

Domain Knowledge and Digital Tools:

Effective implementation of AI/ML requires a deep understanding of the energy market. Domain expertise is crucial for selecting appropriate models, interpreting results, and translating insights into actionable strategies. Collaboration between business and technology teams is essential, fostering a culture of data-driven decision-making.

 

The Path Forward: A Smart Energy Future

The energy sector is at a pivotal juncture, driven by increasing demand, market volatility, and the urgency for sustainability. Embracing AI and advanced analytics is essential for organizations seeking to thrive in this dynamic environment. These technologies are essential for navigating the complexities of integrated supply chains, characterized by fluctuating demands, diverse supply sources, and logistical constraints.

By leveraging AI and ML, organizations can process vast datasets, build sophisticated predictive models, and continuously refine them for optimal decision-making. Successful implementation requires not only technological adoption but also close collaboration between business and technology teams, ensuring domain expertise aligns with advanced analytical capabilities. This empowers organizations to optimize operations, mitigate risks, and drive innovation, ultimately fostering a resilient and sustainable energy future.

 

 

Artificial Intelligence with Human Engagement – a powerful combination for the digital but humane future

 Artificial Intelligence is creating waves as the world gets digitized, processes get automated in the face of rising global population, concerns about environment and managing resources and livelihoods across the world. World population has just crossed 8 billion and stands at 8.1 billion, looking to reach 10.3 billion by 2100. Many of the problems with that growth can be fixed with technology.

Technology has put the world in the palm of one’s hands. AI is a great innovation with amazing features and potential to really impact the world, positively. It can impact human lives and livelihoods, across manufacturing, medical sciences, construction, energy, environment and so much more. It’s a fantastic tool when used with conscience and intelligence; a dangerous weapon if not used properly.

BACKGROUND

AI or Artificial Intelligence has a long history, unknown to many who see this as a relatively modern phenomenon. The field of AI research was founded during a summer conference at Dartmouth College in the mid-1950s, where John McCarthy, computer, and cognitive scientist, coined the term “artificial intelligence. Marvin Minsky (Carnegie-Mellon University) defines AI as "the construction of computer programs that engage in tasks that are currently more satisfactorily performed by human beings because they require high-level mental processes such as: perceptual learning, memory organization and critical reasoning”. 

Why has AI taken off now?

Artificial Intelligence has started disrupting a wide variety of domains. There are a few reasons AI is taking off now.

·       Availability & Access to Data: Data is constantly being shared by users and applications in public and private domains. Previously, customers and businesses were skeptical about sharing data and often held onto the data for fear of leaking information. Now, data is being shared continuously, knowingly, and mostly, unknowingly e.g., location, your preferences, topics you are interested in, items you like to buy, and activities you are doing and interested in. And, all this data is available in a digital format, ripe for computers and machines to consume. Some of the information is available in textual or pictorial form and may need significant processing to extract actionable data.

·       Data storage: With the advent of new technology and cloud infrastructure, data storage has become cheap and capacity almost limitless. Huge sets of data are now stored and readily available in structured or unstructured form.

·       Computational ability: Processing power to gather, clean, analyze, and perform computation of the data has increased exponentially. Big data sets are processed, analyzed, trained, and inferred from easily, much quicker than earlier. Issues around consumption of data and processing them efficiently and quickly do exist and more research is being done.

·       Sophisticated algorithms: With decades of research, algorithms have become smarter and mature. Advancements in Data Sciences, Machine Learning, Heuristics, neural networks, and recently Quantum Computing, has made processing and derivation of results much faster and more relevant.

Artificial Intelligence – it’s wide range of uses

Artificial Intelligence (AI) has a wide range of applications across various fields. Here are ten of its top uses:

 

·       Healthcare: AI is used for medical image analysis, disease diagnosis, personalized treatment plans, drug discovery, and patient management, improving accuracy and efficiency in healthcare delivery.

·       Finance: AI is employed for algorithmic trading, fraud detection, credit risk assessment, customer service through chatbots, portfolio management, enhancing decision-making and risk management in the financial sector.

·       Autonomous Vehicles: AI powers self-driving cars, enabling them to navigate, detect obstacles, and make real-time decisions, potentially revolutionizing transportation and reducing accidents.

·       Natural Language Processing (NLP): NLP techniques are used in chatbots, virtual assistants, language translation, sentiment analysis, content generation, and more, facilitating human-computer communication and language-related tasks.

·       E-commerce and Marketing: AI aids in personalized product recommendations, customer segmentation, demand forecasting, and targeted advertising, leading to improved customer experiences and better marketing strategies.

·       Manufacturing and Industry: AI-driven robotics, automation, predictive maintenance, and quality control optimize manufacturing processes, leading to increased efficiency, reduced downtime, and enhanced product quality.

·       Education: AI can significantly impact the education domain by becoming the ideal personal tutor and having a significant influence on average kids. It will figure out the gaps, focus on addressing them and customizing instructions for the student.

·       Entertainment: AI is employed in video game development, content recommendation systems for streaming platforms, music composition, and special effects, enhancing entertainment experiences.

·       Cybersecurity: AI helps in detecting and responding to cyber threats by analyzing patterns, identifying anomalies, and predicting potential attacks, bolstering the security of digital systems and networks.

·       Agriculture: AI assists in crop monitoring, disease detection, yield prediction, and precision farming, optimizing resource utilization and increasing agricultural productivity.

·       Energy Management: AI is used in optimizing energy consumption, grid management, predictive maintenance of energy infrastructure, and renewable energy generation, contributing to more sustainable and efficient energy systems.

·       Computer Programming: Just by describing in plain English the purpose of the program, tools like CoPilot can quickly generate standard code in any programming language and thereby free up the programmer from writing run-of-the-mill programs and enable them to develop more complex and niche programs. 

 

These are just a few examples, and the applications of AI continue to expand across various industries as the technology evolves.

Human involvement as AI grows.

As Artificial Intelligence impacts every digital interaction of our lives and livelihoods, humans will play a critical role on many fronts. It is essential to understand that AI still needs the expertise of humans to finally pull the trigger. Even though things are going digital, and data driven, the decisions being made are for you or on your behalf, so they continue to be guard railed by the norms in that industry or society or professions.

·       Expertise – AI is based on inference engines and knowledge bases. Human expertise in defining the necessary and sufficient rules and constraints, developed empirically and intuitively by experts over decades forms the foundation of the inference engine, and decision making.

·       Efficiency - Technology will automate, remove the boring, painful manual jobs.AI will help humans to focus on more value-added fulfilling jobs, which leverage the higher skillset, more challenging and satisfying. Mundane work takes up unnecessary time; humans can deliver much more value focusing on high-skilled “brainier” work.

·       Involvement – Human involvement, learning from actions taken, helping with decision making are key to the success of AI. These are jobs which often lead to manual errors, which lead to more effort in reconciliation and fixing the errors.

o   Many of the uses highlighted above help businesses and people in processing humongous amount of data, run many complex models, weed out scenarios and help in decision making and taking the right action.

o   AI will help humans to do things like review results or scans and diagnose diseases or potential problems in the future. Diagnosis which would have been difficult for doctors based on just his knowledge, can be possible as computers leverage AI, much larger datasets, and rules.

o   Key is they use human’s expertise and knowledge to do things which was impossible for humans to do efficiently and then, deliver the results for humans to act on.

 

·       Bias – It’s critical to train AI models to eliminate obvious biases tied to individuals. Biases that exist in current algorithms and models can be better deciphered and rectified by a wider group involvement and continuous reviews of processes.

·       Validation – Humans are needed to validate the data and ensure the integrity of the data. They need to validate the models and the results and determine them to be suitable for action.

·       Decision making – AI will be a great tool to augment human decision-making. For example, an expert geoscientist, looking at 100 spreadsheets to decide on which field to explore, can now rely on AI and automation to process a million data sources, run complex models, and give the top 5 scenarios to consider. Humans still make the final decision, but it is based on many more data sets, scenarios, models, and complexities. AI will not make the final decision; humans will in many cases. For this balance to work, it is critical to know where AI is used and how any potential bias can be addressed.

·       Ethics and Empathy – One of the biggest concerns is AI going berserk without any human intervention, no consideration for human empathy, softer side of jobs, services, and interaction. Already significant investment is happening to put humans as arbitrators and, to research and model human features into AI models. AI cannot be an uncontrolled Frankenstein; it should be a better version of what humans could do alone but now can be achieved by working with humans. Humans can apply these softer skills, subjectively judge between options, and make the best use of AI. There should be a strong governance angle to this – where AI and its applications need to be monitored and proper oversight needs to be provided, like Asimov’s law of robots or Microsoft’s AI standards.

·       Augmented Intelligence – Finally, the combination of human and artificial intelligence created “Augmented Intelligence”. They close each other’s gaps; while AI processes more data, more models, and performs tasks automatically that humans cannot; human intelligence brings in empathy, subjectivity, validation, and many of the subjective mental inferences, machines may not be capable of.

 

Conclusion

Human intervention and natural intelligence must work together with Artificial Intelligence to bring the biggest value to humanity. Machines cannot be left to their uncontrolled will; humanity, empathy, and governance should be in place to monitor, control and pick the right uses of AI. That will impact the world, positively and for the greater good of humanity.

 

 

 

 

 

 

 

Driving digital strategy for the rapidly changing energy future

As the world embraces a new energy future, digital transformation of the energy industry has become critical. Energy future will be defined by ability to handle political and economic crisis, incredible growth in energy needs, value for green sustainable energy including cleaner part of the carbon spectrum, ability to handle disruptions in production and transportation, and last but not the least, ability to accurately forecast demands and prices and serve customers better.

Digital Transformation is a business-led initiative with IT as partner

Such a complex set of factors calls for increased use of sophisticated digital technology to gather, process and analyze data faster and smarter to drive automation, improve decision making and better cater to customers and counterparties. The problem in that scenarios is not to hunt for new and more sophisticated technology but to understand the domain, absorb the usecase and need and better design the right solution, leveraging the right technology.

Hence, driving digital strategy will increasingly involve having the right understanding of the business domain, its use usecases as also understanding of the technology options to devise the right path forward. Getting the digital strategy right will be critical to the success of any company in the future, successfully predicting and addressing customer and counterparty needs and getting a leg up on the market, running operations efficiently.

In short, digital strategy will define a company to its customers, counterparties and partners. Company should have that as a clear high priority for success. This calls for a digital ambassador to partner with business. Digital transformation is NOT just an IT initiative – the foundation of digital transformation is a close partnership between business and IT, driven by business goals.

Data driving customer centricity

Every company under the planet can potentially leverage digital technology to gather more information about what the customer needs, what type of goods or services they want, gaps and positive feedback and many other critical information. Key component of the digital strategy must involve gathering data to value customer needs better and continue monitor customer engagement and feedback to make their experiences better.

Data and Analytics leading to better decision making

Business world is now much more integrated with a global supply chain with challenges in one area potentially impacting the entire chain. It is imperative to have sophisticated models evaluating process, demand, supply, logistics and transportation in analyzing disparate sources of data and making optimal decisions. A sound digital strategy should involve integrated data platform and advanced analytics tools, including AI / ML.

Automation driving efficiency

Key feature of the business process landscape is the exposition of processes all through the day. That has opened a universe of opportunities to automate them and save human effort on more valuable tasks, requiring human intervention. A digital strategy should look into processes and tasks, which can be automated – from gathering data, creating reports, triggering jobs, monitoring alerts etc. Digital Twin is now helping organizations with running, maintaining, monitoring processes, evaluating simulations and acting as a digital assistant to an onsite expert. These automations will drive efficiency across the organization.

Conclusion

Technological advances like cloud computing, data and analytics, AI/ML, automation/RPA, blockchain and others have significantly impacted the solutions landscape. There is a plethora of tools available to address various problems. It is critical to have a business strategy, starting with a clear understanding of business goals and knowledge of the business domain. That should enable one to devise the right strategy and technical solutions to help deliver value for business.

Having a digital ambassador to clearly articulate the value of digital technology to business and then partnering with business in delivering solutions will be key to successful digital transformation of the energy industry. With a solid partnership between business and IT, delivering right solutions for the right business problems in an agile way, digital transformation will indeed change the organizational landscape and help deliver tangible and consistent results.

Future of energy – where energy transition and digital transformation converge

 

Future of Energy is evolving at a rapid pace – where soaring global energy demand, increased volatility as also energy transition and digital transformation converge.

In that context, I attended and spoke at the Energy Risk USA event on May 11, 2023, the annual gathering of experts across the energy industry, involved in trading, supply and risk management of energy commodities. It spans all the commodities – oil, gas, refined products, natural gas, LNG, and power, in all forms including renewables.

The summit had great relevant topics, thoughtfully designed. There was top notch engaging discussion and presentation and such a fantastic interaction with a diverse group of experts on the future of energy, from energy transition to digital transformation.

 

Key highlights across sessions:

·       Energy demand - With more gadgets, increased use of technology, energy demand is supposed to rise, across the globe.

·       Market prices/volatility – Expect higher volatility as economic/political factors, supply constraints, interconnected global markets, energy transition mingle.

·       Credit Risk/Margin – As a result of increased volatility, there may be requirement for higher margin and collateral. Credit agency and regulators will have to work hand in hand as the energy future evolves and need to invest in new energy projects rises.

·       Environment/ Clean energy – There is significant focus on environment and clean energy. That is driving not only interest in renewables like wind, solar, hydrogen, biodiesel and geothermal but also in cleaner spectrum of carbon fuel – natural gas and LNG.

o   There are significant constraints in energy storage to handle hours when renewable will not be prepared to produce power. This will create more volatility even though battery storage is somewhat of a hedge.

·       Balancing between various forms of energy – In order to meet increased demand for electricity specially when renewable energy is not available, it is critical to balance traditional sources of energy with the renewable sources.

·       Investment in new energy – New energy investment, calculating Risk Adjusted Rate of Return, managing capital need to be sorted out. New models handling future investment specially in renewables must be developed.

·       Modelling and Quantitative tools - There will be need for sophisticated modelling to properly track and quantify risk and margin. It will be needed to forecast price, demand, supply and properly address demand in the most efficient way. There will be need to manage risk, margin and working capital as volatility increases.

·       Data and Analytics – There will be need for leveraging more and diverse types of data, running analytics and help make better decision. Usage of cloud technology and advanced technology will be a key for energy future.

·       Artificial Intelligence, Machine Learning, and tools like ChatGPT

o   As world moves to an integrated supply chain, with more volatility related to rising energy demands and different supply curves based on energy source, more logistical constraint, there will be increased demand to leverage advanced technology to solve problems, make better decisions. AI and ML will be used to handle and model huge amount of data, leveraging AI running sophisticated models, using Machine Learning to tweak models and help better outputs.

o   In addition to using structured data and numbers, it was important to extract value from unstructured data, existing information in the public domain. ChatGPT can be leveraged as a natural language processing tool to mine huge amount of public data, even unstructured data like social media comments to help model the future and do analytics. E.g., People talking about big events, incoming demand in Asia can be mined to help develop forecast for demand. Again, this will be in addition to the quantitative models and predictive analytics.

·       Domain knowledge and Digital tools – It was universally clear that it was not enough to just deploy the smartest and coolest technology. Domain knowledge, understanding the business problem to pick the right model and right tool was critical.

o   It means there has to be close collaboration between business and technology; importance to have data scientists with domain knowledge, business experts with access to digital tools.

Energy Risk USA provided the platform for deep discussion around not only energy, environment, market factors but also how to automate, digitalize the future as resources become scare, logistics become complex and demand soars. It is an exciting time to plan for the smart and green energy future.