Discover how 60+ global companies are leveraging AI
Sector
Company
Topic
AI Category
Impact
Summary
Read More
Number of case studies
Consumer Discretionary
Gap Inc.
Gap Inc. Elevates online shopping with AI
Generative AI
Improved user experience
Gap Inc. is elevating the online shopping experience by embedding AI directly into the buying journey, pairing personalized fit guidance with seamless conversational checkout to remove two of the biggest friction points in digital apparel shopping. Through Bold Metrics’ predictive sizing technology, customers receive real‑time, conversational size recommendations instead of static charts, helping them choose the right fit with confidence at the moment of purchase. At the same time, Gap Inc.’s support for Google’s Universal Commerce Protocol (UCP) ensures its products appear accurately and are instantly purchasable across AI‑native environments like Google Search and the Gemini app, enabling smooth, agent‑driven checkout wherever inspiration strikes.
1
Consumer Discretionary
Mercedes‑Benz Group
Mercedes-Benz Direct Chat
Generative AI
Time saving
Mercedes‑Benz Direct Chat is an internal, AI‑powered assistant that streamlines daily work by giving employees fast, context‑aware support across writing, research, and collaboration tasks. Built on ChatGPT and enhanced in 2025 with Google Gemini integration, the tool now lets teams switch between multiple AI models and even generate or analyse images through DALL·E, expanding its creative and analytical capabilities. Direct Chat also connects to Mercedes‑specific knowledge agents, turning it into a central hub for internal information and cross‑functional workflows. With more than 10,000 daily power users and availability in both English and German, it has become a core productivity tool across the organisation. Importantly, the system is designed with strict data‑protection measures—anonymous inputs, Azure OpenAI’s secure infrastructure, and European data‑centre hosting—ensuring responsible, compliant use of generative AI at scale.
1
Consumer Discretionary
Mercedes‑Benz Group
MBUX features context-sensitive awareness with AI
Machine Learning
Improved user experience
Mercedes‑Benz’s MBUX system delivers personalized, proactive suggestions by using AI to understand each driver’s habits, context, and preferences, reducing the number of steps needed to access key functions. The system continuously learns from behavior and surroundings to surface the right feature at the right moment—from navigation and media to comfort settings—directly on the “zero‑layer” interface, eliminating the need to dig through menus. Examples include suggesting a weekly call you usually make on Tuesday evenings, recommending the hot‑stone massage function when temperatures drop in winter, prompting seat‑heating‑related comfort features when you activate seat heating, or automatically offering to raise the chassis when approaching a location where you previously used the lift‑up function. These “Magic Modules” turn MBUX into an intelligent co‑pilot that anticipates needs, personalizes the cabin experience, and makes in‑car interaction more intuitive and effortless.
1
Energy
Adnoc
'ENERGYai' - Agentic AI tailored for the Energy Sector
Agentic AI
Improved operational efficiency
ENERGYai – is an agentic AI solution, tailored for the energy sector. Developed in the UAE with AI powerhouse AIQ in collaboration with Microsoft and G42, ENERGYai uses AI agents combined with large language model technology, built on decades of proprietary ADNOC knowledge and trained on workflows across ADNOC’s value chain.
ENERGYai brings a new level of efficiency and precision to critical tasks from seismic analysis to geological modeling and real-time process monitoring, reducing time for essential business processes from months to days, minimizing cost and emissions in the process.
1
Energy
Aker BP
'Exploration Robot' - can Gen AI help find Oil
Generative AI
Improved workflow
Aker BP has developed a Generative AI based “exploration robot” that accelerates and enhances the process of identifying oil and gas prospects by analyzing vast, multi‑decade seismic, well‑log, and core‑sample datasets far beyond human cognitive capacity.
Rather than replacing geologists and geophysicists, the system acts as a co‑pilot—flagging potential targets, quantifying uncertainty, and automating repetitive interpretation tasks so experts can focus on higher‑value creative and analytical work.
Cordant is a modular, AI-enabled industrial enterprise solution designed to optimize assets, processes, and energy use at scale. Cordant combines world-class sensing and protection hardware with open digital solutions to transform raw data into actionable insights that unlock new levels of operational efficiency across the enterprise.With advanced Al and machine learning capabilities and seamless integration with existing operations management and enterprise systems, Cordant acts as a digital thread across operations, connecting data, automating decision-making, and delivering predictive insights for improved asset reliability, process efficiency, and sustainability. By moving beyond siloed, plant-level tools, Cordant elevates the value of operational data to the enterprise level, enabling strategic, organization-wide impact, achieving business goals, hitting key performance indicators, and improving operational efficiency ratios.
1
Energy
Chevron
Chevron’s 'ApEX' - AI solution helping Oil and Gas discovery
Generative AI
More competitive bids
Chevron’s ApEX is a proprietary generative‑AI platform that transforms how the company prospects for oil and gas by rapidly analyzing more than a million exploration files to deliver real‑time insights. Launched in 2024, ApEX uses multiple specialized search agents—each focused on areas like geospatial data or prospect evaluation—to retrieve, synthesize, and contextualize subsurface information that previously required extensive manual work.By improving access to hard‑to‑find data and surfacing overlooked seismic interpretations, logs, and technical reports, ApEX helps Chevron make faster, better‑informed exploration decisions, strengthening competitiveness in high‑stakes acreage bidding and accelerating the path from idea to actionable prospect.
1
Energy
Chevron
'APOLO' helps Chevron pinpoint prime Drilling locations
Machine Learning
Better informed decisions
Chevron’s APOLO platform is an AI‑driven system that analyzes millions of subsurface and production data points from shale and tight assets to generate fast, standardized, and highly accurate well‑performance forecasts, helping engineers pinpoint optimal drilling locations and design more productive wells. Unlike traditional generalized models, APOLO captures local geological variations and provides explainable insights, enabling smarter decisions about spacing, proppant, and fluid use. It continuously learns, simulates alternative well designs, and accelerates development planning—ultimately improving capital efficiency, strengthening portfolio resilience, and supporting Chevron’s long‑term strategy to optimize profitability across the Permian, DJ Basin, and future global assets.
1
Energy
Chevron
Robotics supports more efficient workplace
Robotics
Improved operational efficiency
Chevron is deploying robotics—including tank‑cleaning robots, inspection crawlers, and autonomous drones—to improve safety, reduce costs, and eliminate high‑risk work across its operations, achieving more than $92 million in savings and removing over 143,000 at‑risk hours since 2024.These systems handle hazardous tasks such as entering confined storage tanks, conducting thermal and visual inspections, and detecting emissions, allowing employees to focus on higher‑value work. The program, demonstrates how robotics can enhance both operational efficiency and workforce safety.
1
Energy
ENI
EnergIA – Eni’s generative artificial intelligence tool
Generative AI
Faster access to information
Eni’s “EnergIA” tool is based on generative artificial intelligence and helps users to rapidly locate information on eni.com.
The system automatically integrates new content, almost as soon as it is published on the website.
1
Energy
Equinor
Machine learning against the machine
Machine Learning
Improved failure detection rate
Omnia.Prevent is a maintenance solution based on machine learning. The journey starts with rotating equipment such as turbines, which have an Internet of Things (IoT) device attached to them. These IoT devices pick up on changes in temperature, pressure and more. The Prevent team have trained machine learning (ML) algorithms to look for changes in this data.
1
Energy
Equinor
Drones and robots in Equinor
Robotics
Improved safety
A project called AIR (Autonomy, interoperability and robotics) has been established to deliver on the part of Equinor’s digital vision that cover robotics: “Robots will make our work easier”. It also delivers on the company’s strategic goals of always safe and high value.
1
Energy
Equinor
Use of AI saved Equinor USD 130 million in 2025
Machine Learning
Cost saving
Equinor reports that artificial intelligence delivered USD 130 million in value creation and cost savings in 2025, driven by large‑scale deployment of AI across offshore platforms and onshore facilities. The company now uses AI to interpret seismic data ten times faster, generate thousands of well‑planning scenarios, and optimize field development—capabilities that even uncovered a novel solution at Johan Sverdrup Phase 3 that saved USD 12 million.Predictive maintenance has been especially impactful, with AI monitoring more than 700 rotating machines via 24,000 sensors, preventing failures and contributing USD 120 million in value since 2020. In total, AI has generated over USD 330 million in realized value since 2020.
1
Energy
Halliburton
Investigating Invisible Lost Time on textual data using AI
NLP
Improved planning
During the well construction process, activities are described and recorded by the onboard crew. This produces a massive amount of textual data in daily rig reports. Later, these records are used to detect specific events and calculate the duration of each operation inside the well to compare with performance metrics. Invisible Lost Time (ILT) is the difference between the actual operation duration and a best practice target (or a technical limit) to understand if a well is performing better or worse from correlated scenarios.To perform ILT analysis, all manual descriptions must be checked by an SME who can correctly identify the type of ongoing operation. However, manual classification consumes a lot of time and may be done by multiple stakeholders. This may lead to inconsistencies caused by their subjective interpretations. To help overcome this problem, an Artificial Intelligence (AI) solution based on Natural Language Processing (NLP) was developed. The solution consolidates multiple data sources and then processes all necessary data to help quantify the impact of ILT.
1
Energy
Halliburton
Using AI to redevelop declining brownfield
Machine Learning
Faster operations
Halliburton’s Malaysia brownfield redevelopment showcases how tightly integrated automation and AI can transform complex well construction into a high‑precision, low‑risk operation. By combining LOGIX automation’s physics‑based machine‑learning predictions with real‑time Geo‑Span® downlink steering and the iCruise® intelligent RSS, the team autonomously steered 97% of the 8½‑inch section and executed 84+ closed‑loop commands flawlessly, even within a congested multi‑well pad and multi‑stacked reservoir environment. The wellbore effectively “painted over” the planned trajectory—just 1.01 m below and 0.24 m left of plan—while achieving a 32% faster ROP than previous campaign wells and saving 14 hours of rig time, equivalent to USD 122,500 in avoided cost. This case demonstrates the operational uplift possible when digital twins, ML‑driven steering, and autonomous drilling workflows converge in the field.
1
Energy
Halliburton
Improving drilling performance with automation and AI
Machine Learning
Faster operations
Halliburton’s case from Colombia shows how integrating LOGIX automation, machine‑learning driven predictive analytics, and the iCruise intelligent RSS materially improved drilling performance in deep, complex horizontal wells. Operating in the geologically challenging Llanos Basin, where seismic faults, regressive‑transgressive formations, collision‑avoidance constraints, and tight KPIs increased operational risk, the operator shifted from traditional descriptive optimization to a fully predictive, AI‑enhanced workflow. Intelligent sensors feeding a digital twin and physics‑based BHA model enabled real‑time trajectory optimization, reduced tortuosity, and closer alignment with planned directional difficulty indices. The unified human–AI team delivered consistent performance gains across three wells, achieving a 33% increase in ROP and 15–45% faster casing and liner running speeds, while maintaining smooth hole profiles and minimizing deviation from plan.
1
Energy
Nabors Industries
Predictive Drilling Solution improves average ROP by 36%
Machine Learning
Time saving
The collaboration between Corva and Nabors has produced a breakthrough Predictive Drilling solution that integrates Corva’s AI/ML ROP optimization models with Nabors’ SmartROS® and RigCLOUD® automation platforms to deliver real‑time, closed‑loop control of rig auto-drillers.
In a Delaware Basin trial, the system increased average ROP by 36% and reduced vibration by nearly 10%, while achieving rapid field adoption—80% of lateral footage drilled on the first pad—thanks to seamless deployment and minimal workflow disruption.
1
Energy
National Oilwell Varco (NOV)
Intelligent Drilling Optimizer - Kaizen
Machine Learning
More efficient drilling
Kaizen Intelligent Drilling Optimizer (IDO) is NOV’s AI‑driven drilling optimization system that continuously learns from real‑time wellbore conditions and offset data to proactively mitigate dysfunction and maximize performance. By evaluating drilling dynamics on the fly and recognizing environmental changes, Kaizen automatically determines optimal weight‑on‑bit and RPM setpoints, operating either in advisory mode—where drillers receive recommendations—or in full control mode, where it sends setpoints directly to the autodriller for closed‑loop execution. Its digital‑twin models and machine‑learning algorithms go beyond traditional MSE‑based optimization, identifying and mitigating stick‑slip, bit bounce, axial and torsional oscillations, while avoiding drillstring natural frequencies and monitoring distributed stress to prevent buckling or overstress. Field results highlight its impact: drillers report significant improvements in ROP, smoother drilling, and more consistent performance—even compared to seasoned human judgment—demonstrating how autonomous, continuously adapting control can materially elevate drilling efficiency and wellbore quality.
1
Energy
Repsol
A 'smart Copilot' to help you at work
Generative AI
Increased productivity
Repsol conducted a large‑scale, four‑month study with 600 employees to evaluate Microsoft 365 Copilot’s impact on everyday work, finding that the generative‑AI assistant saved participants an average of 121 minutes per week by accelerating tasks such as information retrieval, document summarization, and drawing conclusions. The experiment, published in MIT Technology Review, showed that Copilot not only improved productivity but also enhanced the quality of outputs—documents produced with the tool scored 16.2% higher—while helping employees overcome creative blocks and navigate complex tasks more intuitively. Participants highlighted the tool’s human‑like interaction style, its usefulness in real business scenarios, and the rapid learning curve that unlocks its benefits, leading two‑thirds of users to say they would not want to return to working without generative AI.
1
Energy
Shell
Solving the 10-year-old problem at Perdido
Machine Learning
Increased operational uptime
Shell’s team of “data detectives” worked on solving a decade‑long mystery on the Perdido deep‑water platform in the Gulf of Mexico: the recurring failure of pumps responsible for separating oil and gas. By combining real‑time sensor data—temperature, pressure, and chemical signatures—with artificial intelligence, the team searched for patterns that could predict disruptions before they happen. After initial attempts fall short, collaboration with an experienced pump operator helped refine the models, ultimately revealing a chemical signature that appears ahead of roughly 70% of failures, marking a major breakthrough in using AI to enhance offshore reliability and operations.
1
Energy
SLB (Schlumberger)
Tela agentic-AI assistant
Agentic AI
Time saving
Tela is an agentic-AI assistant that learns, reasons, and acts autonomously,
with a unique understanding of the energy industry.
Each Tela agent understands the nuances of subsurface, drilling, production, and sustainability workflows. It can work interactively with the user and can also work proactively in the background,
monitoring context, coordinating tasks, and triggering actions without constant user input. This means faster
decisions, fewer errors, and workflows that adapt to user needs.Key capabilities at a glance:
- Domain-aware agents for subsurface, drilling, production, sustainability, and more.
- Proactive automation that anticipates needs and streamlines routine and complex tasks.
- Continuous reasoning for monitoring, validation, and decision support.
- Cross-workflow coordination that connects tasks and systems seamlessly.
1
Energy
Woodside Energy
'Maint Intel' - Right fix at right time
Machine Learning
Optimized asset management
Maint Intel is an AI-enabled tool that helps optimize the frequency of equipment maintenance, ensuring preventative work is scheduled when it’s needed.
It recommends optimal maintenance schedules by analyzing huge volumes of data and comparing maintenance plans and historical performance against reliability targets.
1
Financials
BBVA
BBVA incorporates Generative AI into their Blue app
Generative AI
Improved user experience
BBVA has updated its personal virtual assistant Blue, integrated in its app, making it more personalized. Blue now has enhanced abilities to interact with customers using natural language, provide tailored information on their finances, and perform some of the most common account and card transactions.
1
Financials
Citigroup Inc.
Citi Stylus Workspaces with Agentic AI
Agentic AI
Increased productivity
Citi rolled out an enhanced version of its proprietary AI platform, Citi Stylus Workspaces, now powered by Agentic AI.
This latest version integrates directly with Citi systems, linking seamlessly to internal data and project management tools, including the company’s global employee directory, commonly used enterprise platforms and other key resources, while also leveraging web searches and analysis capabilities.With this update, employees can conduct in-depth research, extract insights from vast datasets and streamline multi-stage workflows into a single, automated process.
1
Financials
Citigroup Inc.
CitiService Agent Assist
Generative AI
Improved user experience
CitiService Agent Assist, a generative AI-powered solution designed to help customer service agents deliver faster, more accurate responses to clients.Agent Assist guides customer service teams through client interactions with procedural information, real-time transcripts, after-call summaries and more, helping save time and improve client outcomes.
1
Financials
DBS Group Holdings Ltd.
Tapping into AI-powered nudges
Machine Learning
Improved user experience
DBS created more than 100 artificial intelligence and machine learning algorithms that analyze an internal data mart with 15,000 customer data points, to generate seven types of nudges that, for instance, offer personalized product recommendations and celebrate customers’ milestones.
1
Financials
HSBC Holdings plc
Transforming HSBC with AI
Generative AI
Improved user experience
HSBC boasts a number of use cases for generative AI including:
- a generative AI assistant for their servicing teams in Corporate and Institutional Banking, that supports 3 million client interactions annually, reducing turnaround times and improving experience.
- to support credit analysis write-ups using trusted internal and external data sources, reducing time to complete the credit application process.
1
Financials
HSBC Holdings plc
Harnessing the power of AI to fight financial crime
Machine Learning
Faster detection rate for financial crimes.
Financial crime doesn’t stand still, the tactics used by fraudsters are constantly changing.HSBC uses AI to help check about 900 million transactions for signs of financial crime each month, across 40 million customer accounts.As new financial crime tactics or trends emerge, HSBC teaches their AI what to look out for. As a result, they’re able to find and tackle financial crime faster.
1
Financials
Interactive Brokers
IBKR’s AI-Powered Commentary Generator
Generative AI
Improved productivity
Commentary Generator leverages generative AI to create client-specific reports summarizing the latest performance and ticker-specific news from select providers. Commentary Generator helps advisors streamline their workflow and improve efficiency.
1
Financials
JPMorgan Chase & Co.
Tech for Social Good partners with Tata STRIVE
Machine Learning
Positive social impact
Tech for Social Good, a JPMorgan initiative, partnered with Tata STRIVE—an Indian skill‑development program for underserved youth—to tackle high student dropout rates. Tata STRIVE needed an early‑warning system to identify participants at risk of leaving the program. Tech for Social Good built machine‑learning models that progressively improved, eventually predicting dropouts with 96% accuracy two weeks in advance. With these insights, Tata STRIVE was able to intervene earlier and ultimately reduced dropout rates among at‑risk learners by 33%.
1
Financials
JPMorgan Chase & Co.
LLM Suite - a Gen AI 'Research Assistant'
Generative AI
Increased productivity
JP Morgan's LLM Suite is a generative AI tool designed to serve as a ‘research analyst’ for employees. The LLM Suite stands out for its ability to generate and refine written documents, provide creative solutions and summarize extensive documents, thus streamlining the information flow within the company. These capabilities not only save time but also improve the quality of output, making the tool invaluable for employees across different departments.
1
Financials
JPMorgan Chase & Co.
The Machines Behind the Market
Generative AI
Improved user experience
J.P. Morgan’s AI Search is a research engine built directly into the J.P. Morgan Markets platform. It is designed to help institutional investors cut through information overload by delivering fast, trustworthy, and context‑aware insights.The tool is powered by a large language model trained exclusively on J.P. Morgan’s research archive.AI Search allows investors to ask questions in plain English—for example, “How is AI impacting the semiconductor industry?”—and instantly receive concise, well‑structured answers grounded in vetted research. It interprets intent, extracts the most relevant passages, and links back to original reports for transparency.
Mastercard’s Agent Pay is a new agentic payments framework designed to embed secure, seamless transactions directly into AI‑driven interactions, enabling both consumers and businesses to delegate purchases to trusted AI agents with confidence. Built on Mastercard’s advanced tokenisation infrastructure, the programme introduces Agentic Tokens that authenticate AI agents, protect credentials, and ensure transparent, controlled payment flows across the commerce ecosystem. By partnering with Microsoft and other leading AI platforms, as well as enablers like IBM and major checkout providers, Mastercard aims to scale agentic commerce with strong safeguards, clear consumer permissions, and end‑to‑end fraud protection. The result is a future where AI agents can not only recommend products or optimise business operations, but also complete transactions securely on behalf of users—bringing personalised, proactive, and trustworthy commerce into everyday experiences.
1
Financials
Mastercard Inc.
Using AI to prevent consumer payment scams
Machine Learning
Improved user experience
Consumer payment scams have surged into one of the fastest‑growing forms of financial crime, with fraudsters exploiting digital channels, social engineering, and real‑time payments to manipulate people into sending money directly thus making these losses extraordinarily hard for banks to detect or reverse. With global scam losses projected to reach $5.25 billion across major markets by 2026 and institutions reporting rising attack volumes, the industry faces a challenge that traditional security tools can’t solve alone. Mastercard’s answer is Consumer Fraud Risk, an AI‑powered solution trained on billions of transaction patterns and years of mule‑account tracing, enabling banks to identify high‑risk payments before funds leave a customer’s account. By combining global data, machine learning, and real‑time risk scoring, Mastercard gives financial institutions a proactive way to stop scams at the source and significantly reduce the flow of stolen money.
1
Financials
Morgan Stanley
Morgan Stanley Launches AskResearchGPT
Generative AI
Increased productivity
AskResearchGPT is an AI assistant based on OpenAI’s ChatGPT technology. The tool lets users extract answers from across the universe of Morgan Stanley’s research — including on stocks, commodities, industry trends and regions — collapsing what could otherwise be the cumbersome task of gleaning insights from the more than 70,000 reports produced annually by the bank.
1
Financials
Morgan Stanley
Morgan Stanley Launches AI@MS Debrief
Generative AI
Improved productivity
The assistant, called Debrief, keeps detailed logs of advisors’ meetings and automatically creates draft emails and summaries of the discussions.
Debrief replaces the note-taking that advisors or junior employees have been doing by hand resulting in notes of better quality and depth.
1
Financials
Royal Bank of Canada
'NOMI' helps manage day to day spending
Machine Learning
Improved user experience
NOMI is RBC’s built‑in intelligent financial assistant designed to help customers stay on top of their day‑to‑day money management by analyzing cash flow, categorizing spending, and delivering personalized insights directly inside the RBC Mobile app. It offers several integrated features—including NOMI Insights for real‑time spending patterns, NOMI Find & Save for automatically setting aside small amounts based on surplus cash, NOMI Budgets for generating tailored budget recommendations, and NOMI Forecast for projecting upcoming payments and deposits over the next week—all aimed at reducing financial stress and helping users make more informed decisions.
1
Financials
Royal Bank of Canada
'Aiden' - AI for Trading
Machine Learning
Improved user experience
Aiden is RBC Capital Markets’ patented, award‑winning electronic trading platform that applies deep reinforcement learning to optimize trade execution in real time, continuously adapting to shifting market conditions to reduce slippage and minimize market impact. Built in collaboration with Borealis AI, Aiden processes hundreds of market inputs through a deep neural network, learns centrally from every order across regions, and uses a robust reward system to refine its decision‑making with each action. Its ability to proactively explore new trading patterns, anticipate volatility, and stay aligned with key benchmarks like VWAP and Arrival Price positions it as a next‑generation alternative to traditional, rule‑based algorithms.
1
Healthcare
Amgen
How AI is used to Support Patient Access & Reimbursement
Generative AI
Faster support for patients
Amgen recently began using an AI agent named Eva developed by Infinitus Systems, Inc. to assist with the benefit verification process. Designed to complement and support human review, Eva identifies itself as an automated system when contacting health insurers to help collect coverage details and generate benefit statements quickly and reliably.
1
Healthcare
Amgen
Follow the Data
Machine Learning
Faster clinical trial enrollment
- Amgen developed ATOMIC, a machine learning (ML) model that analyzes large amounts of data and predicts which clinical trial sites are likely to enroll patients more quickly and effectively.
- ATOMIC leverages many forms of data, including demographic, geographic and real-world data to ensure faster and more representative participant enrollment.
- An internal analysis of 13 Amgen-sponsored studies showed enrollment was on average three times faster at highly ranked ATOMIC sites than lower ranked sites.
1
Healthcare
AstraZeneca
How AI and Machine Learning augments Drug Discovery
Machine Learning
More effective drug discovery
Machine learning is improving how AstraZeneca generate, screen and evaluate molecules as potential candidate medicines.The number of possible drug-like chemicals is vast. Developing new medicines therefore requires researchers to consider very large numbers of potential molecules (the ‘molecular space’) to identify suitable candidates that can then be investigated in more depth for their utility as potential therapeutics.To help narrow the search for lead candidate molecules, AstraZeneca paired graph neural networks with ‘transfer learning. These processes start with a very large number of molecules and progressively ‘funnel’ the focus towards an ever-smaller number of molecules with the desired medicinal effects and safety profiles.AstraZeneca used transfer learning to store the knowledge from datasets that are large and easily generated at the early stages of the drug discovery funnel (but, on their own, provide limited insights) to improve the predictive performance at the later stages of the funnel – where datasets are more expensive to generate but can provide deeper insights.
1
Healthcare
AstraZeneca
Using advanced Molecular Imaging in Drug Discovery
Machine Learning
Improve drug discovery
Molecular imaging is an advanced form of imaging that enhances our understanding of the effects of drug compounds on human tissues at a cellular level.Mass spectrometry imaging (MSI) stands out as a powerful type of molecular imaging, allowing us to better understand the human body’s molecular complexity and what exactly a candidate medicine is doing in a patient’s body. Using this technology, we are able to measure the individual masses of molecules – whether peptides, proteins, lipids, endogenous metabolites or drug molecules – using a mass spectrometer and simultaneously visualize their spatial distributions. Learnings from MSI offer vital clues to understand the inter-relationships of these molecules within the tissue microenvironment and allow us to better assess the safety and efficacy of our medicines.The whole sample is scanned a few microns at a time, providing a wealth of digital information. From this, we can generate vast datasets from healthy, diseased and drug-treated tissue samples and use AI and machine learning techniques to spot patterns, connections and relationships, turning information into insights and insights into knowledge.
1
Healthcare
Bristol‑Myers Squibb
Predicting new possibilities in drug discovery
Machine Learning
Faster drug development
Bristol Myers Squibb implemented a “predict‑first” model of drug discovery, using AI, machine learning, and massive computational power to forecast which molecules are most likely to succeed long before they reach the lab.
By combining high‑quality datasets with computational tools, researchers can design, evaluate, and optimize molecules in silico, eliminating weak candidates early and accelerating the path to high‑quality therapeutics. This approach is reinforced by “collaborative hybrid intelligence,” where computational scientists and biologists work together to match modalities to mechanisms of action, extract mechanistic insights from complex human biology, and make faster, more confident R&D decisions—ultimately increasing productivity, reducing timelines, and improving the probability of success across the discovery pipeline.
1
Healthcare
Bristol‑Myers Squibb
Mosaic - a new era for tech-accelerated
Patient Care
Generative AI
Improved user experience
Mosaic is a Bristol Myers Squibb’s AI‑powered, end‑to‑end content hub designed to accelerate patient‑centric communication by enabling real‑time identification of healthcare professionals’ educational needs and rapidly generating tailored, high‑quality content at scale.
Mosaic brings together creatives, technologists, and commercialization teams, to streamline content development and enhance personalization for clinicians.
1
Healthcare
Elevance Health
How Gen AI Helps Improve the Consumer Experience
Generative AI
Improved user experience
Virtual Assistant aims to simplify complex healthcare tasks to help people find the information and care they need as quickly and easily as possible. The AI-powered conversational tool can show members whether a medical visit or procedure is covered by their plan, where they can get care, and how much they can expect to pay.
1
Healthcare
IQVIA
Accelerate insights with IQVIA AI Assistant
Generative AI
Improved user experience
IQVIA AI Assistant uses generative AI to enhance users' data-driven decision-making and revolutionize insight generation. It transforms questions and data queries, from simple to complex, into rapid, relevant, and precise answers so users can get the information they need to make decisions with confidence.
1
Healthcare
IQVIA
IQVIA Medical Reasoning LLM (Med-R1 8B)
Generative AI
Enhanced decision making
Med-R1 8B quickly interprets complex real-world data such as clinical notes and scientific literature and provides medically reasoned insights that are easy to understand, enabling experts to make better decisions.Key features include:
- Transparent Decision Chains: Med-R1 8B explains its thought process step-by-step, making it easier for clinicians to interpret, validate, or challenge AI-driven recommendations (see image below).
- Evaluation of Alternatives: Rather than defaulting to a single “best” answer, the model weighs multiple diagnoses or treatment paths before offering conclusions.
- Uncertainty Awareness: Med-R1 8B flags ambiguity, expresses confidence levels, and identifies what information is missing—mirroring how clinicians reason under uncertainty.
- Ease of deployment and cost efficiency: The model’s compact footprint makes it ideal for enterprise-scale use, without the cost, complexity, or infrastructure demands of larger LLMs. It’s easy to integrate, fine-tune, and run securely within clinical environments.
1
Healthcare
Merck & Co.
Accelerating the discovery of new medicines
Machine Learning
Faster discovery of therapeutic candidates
Drug discovery remains an endeavor where only about 1 in 10 drug candidates that enter clinical trials ultimately receive regulatory approval. Merck is working to change that by enabling scientists to use AI and machine learning (ML) foundation models to enhance and build upon their existing approaches to drug design before experimental testing and clinical trials.They developed two foundation models which uncover patterns in disease to find better drug targets, allow faster molecular design and rapidly test small molecules, including cyclic peptides, for efficacy and toxicity before going into the clinic.By unlocking patterns within vast datasets, the AI models enable their scientists to accelerate the discovery of leading therapeutic candidates - a process that normally takes 10 years - allowing them to potentially get therapies to patients faster without compromising scientific rigor.
1
Healthcare
Merck & Co.
Optimizing clinical trials
Machine Learning
Optimized clinical trials
Enrolling people in clinical trials and keeping them engaged once they’ve signed up remains a significant challenge across the pharmaceuticals industry, with approximately 20% of activated sites failing to enroll a single participant. Merck is addressing this by using AI to help improve site selection, patient matching and retention. For example, predictive models can flag patients at higher risk of dropping out, enabling targeted interventions that improve retention and keep trials on track.
1
Healthcare
Merck & Co.
Manufacturing - Gen AI helps protect Supply Chain
Generative AI
Improved risk management
Generative AI helps protect Merck's supply chain when potentially disruptive events like natural disasters or port delays occur. Their system can produce event-based risk assessments in under 30 minutes — allowing them to quickly see which products and sites are affected and act to avoid or reduce shortages and delays.
1
Healthcare
Merck & Co.
Manufacturing - Computer Vision aids Inspection
Computer Vision
More efficient manufacturing
Merck is using computer vision in vaccine manufacturing to inspect vials and syringes for defects. This results in less waste, lower costs and higher production speed.
1
Healthcare
Pfizer
Diversity in Our Clinical Trials | Pfizer
Generative AI
Time saving
Normally, when a clinical trial or trial phase ends, it can take more than 30 days for the patient data to be “cleaned up,” so scientists can then analyze the results. This process involves data scientists manually inspecting the data sets to check for coding errors and other inconsistencies that naturally occur when collecting tens of millions of data points. Thanks to process and technology optimizations, including a new machine learning tool known as Smart Data Query (SDQ), the COVID-19 vaccine clinical trial data was ready to be reviewed a mere 22 hours after meeting the primary efficacy case counts.
1
Healthcare
Royal Philips
AI in Cardiovascular Ultrasound for better diagnoses
Machine Learning
Faster workflows
Philips integrated a new generation of FDA‑cleared AI tools into their cardiovascular ultrasound systems to help clinicians diagnose cardiac conditions more quickly and consistently.
This technology automates tasks such as segmental wall‑motion scoring and 3D quantification of mitral regurgitation to improve accuracy and reduce workload. Trained on anonymized patient data sets from real-life clinical environments, the AI features also help improve the quality and reproducibility of cardiac imaging and enhance operator and departmental efficiency.
1
Healthcare
Royal Philips
'CT Precise Image' - AI aided CT Reconstruction
Machine Learning
Improved workflows
CT Precise Image is Philips’ advanced CT reconstruction technology that uses supervised deep‑learning models to reproduce the clarity and texture of high‑dose filtered back‑projection images while simultaneously reducing radiation dose, lowering noise, and improving low‑contrast detectability, all with reconstruction speeds suited for routine clinical workflow.CT Precise Image was validated by internal specialists and external board‑certified radiologists across a wide range of anatomies and patient types, with assessments focused on sharpness, noise, texture, and artifact reduction. Clinicians retain full control over image generation, with the ability to adjust the strength of Precise Image to match patient needs and clinical judgment.
1
Healthcare
Royal Philips
'CT Precise Position' - Camera-based solution for CT Exams
Machine Learning
Improved workflow
CT Precise Position is Philips’ camera‑based patient‑positioning technology for CT that uses convolutional neural networks to identify anatomical landmarks and automatically align the patient at isocenter, improving vertical accuracy, reducing operator‑to‑operator variability, and shortening setup time. The system analyzes the patient’s body position on the CT table and adjusts placement to support more consistent imaging. Its algorithm has been validated through both clinical and non‑clinical testing using human body phantoms and volunteer studies by Philips’ global clinical teams. Clinicians maintain full control over final positioning decisions, with the ability to review and modify the automated alignment as needed.
1
Healthcare
Royal Philips
'SmartSpeed' - AI-based Image Reconstruction
Machine Learning
Improved workflow
SmartSpeed is Philips’ deep‑learning–enhanced MR acceleration technology that applies an Adaptive‑CS‑Net algorithm directly to the raw MR signal to deliver faster scans and higher‑resolution images across nearly all routine clinical protocols. It builds on the company’s Compressed SENSE platform further enhancing the ability to reconstruct a full image from under-sampled data while maintaining virtually equivalent image quality.The neural network was trained across a wide range of contrasts and acceleration factors to ensure signal fidelity and data consistency.
1
Healthcare
Sanofi
Digital “Twinning”: Clinical Trials Powered by AI
Machine Learning
Faster drug development
Sanofi’s digital‑twinning initiative uses AI‑generated virtual patient populations to simulate how new drug candidates behave long before large human trials begin, allowing researchers to predict safety, efficacy, dosing, and comparative performance with far greater speed and precision than traditional methods.By integrating deep biological data, disease‑pathway models, and quantitative systems pharmacology into a unified computational framework, Sanofi can “test‑drive” investigational compounds in realistic virtual cohorts, validate mechanisms of action, explore treatment scenarios that are impossible or unethical in humans, and even overcome recruitment barriers in rare diseases.These virtual trials reduce early‑stage uncertainty, help prioritize the most promising assets, compress development timelines by skipping certain dose‑finding steps, and continuously improve as more clinical data enrich the company’s growing “scientific memory” of each disease, ultimately aiming toward a future where a broad biology foundation model can predict clinical outcomes across complex therapeutic areas.
1
Healthcare
Sanofi
'Modulus' - Accelerating Manufacturing Agility
Robotics
More efficient manufacturing
Sanofi’s Modulus facility is described as a fully digital, AI‑enabled “factory of the future” designed to transform global biomanufacturing by making production faster, more flexible, and more resilient.
Modulus can simultaneously produce up to four different vaccine or biologic modalities—including viral vectors, mRNA, and recombinant proteins—using a modular, plug‑and‑play architecture that allows technologies to be swapped in and out like components of a Lego set. This flexibility is paired with end‑to‑end automation, robotics, and continuous data capture, enabling real‑time optimization, predictive insights, and rapid adaptation to public‑health demands.
By tightly integrating R&D with Manufacturing & Supply, Modulus also accelerates tech transfer, stability modeling, and launch readiness, enabling delivery of new therapies to patients with unprecedented speed and reliability.
1
Healthcare
Sanofi
'Plai' - an Agentic AI colleague
Agentic AI
Improved input quality across the R&D value chain
Plai is Sanofi’s agentic AI system that sits at the center of portfolio decision‑making, synthesizing more than a billion internal data points to forecast R&D costs, enrollment timelines, and each program’s probability of success, ultimately guiding which drug candidates to accelerate, adjust, or stop.
Acting like a “data‑driven colleague,” Plai provides real‑time what‑if scenarios during governance meetings, flags enrollment risks early, and recommends portfolio optimizations based on unmet need, competitive intensity, and likelihood of clinical success.
Its company‑wide accessibility democratizes insights and elevates strategic discussions by automating data validation so teams can focus on forward‑looking decisions. Plai complements human judgement, driving cultural change, improving input quality across the R&D value chain, and giving Sanofi objective, faster, and more transparent portfolio management
1
Healthcare
Siemens Healthineers
Deep Resolve - unrivaled speed in MRI
Machine Learning
Improved productivity
Deep Resolve is Siemens Healthineers’ AI‑driven MRI acceleration technology that combines deep‑learning–based noise reduction, high resolution reconstruction, and compatibility with parallel imaging to deliver dramatically faster scans (up to 73% quicker) while producing sharper, higher SNR images that outperform classical denoising and interpolation techniques. Deep Resolve ultimately improves clinical productivity while also reducing patient motion and discomfort.
1
Healthcare
Siemens Healthineers
Optiq AI - AI-based live image denoising
Machine Learning
Improved productivity
Optiq AI is Siemens Healthineers’ new AI‑powered imaging chain designed to deliver sharper, lower‑dose images for increasingly complex image‑guided procedures by using real‑time, algorithmic noise reduction and automated exposure optimization across fluoroscopy, acquisitions, and digital subtraction angiography.
This technology helps clinicians maintain high image quality even in challenging scenarios such as steep angulations, or early‑stage minimally invasive therapies, while also accelerating workflows and supporting precision‑focused interventional platforms.
1
Industrials
Carrier Global
Carrier Introduces Generative AI Feature in Abound
Generative AI
Adds clarity to technical responses
Carrier Abound introduced “Tell Me More”, a new generative AI–powered feature within its Abound™ Insights Assistant application that provides additional context and actionable guidance for building operations teams.
When an issue is identified, predictive insights and recommended actions are displayed on the app. The “Tell Me More” feature adds context to reduce ambiguity, support faster resolution and help improve equipment uptime and energy efficiency.
1
Industrials
Honeywell International
Using AI to evaluate Indoor Air Quality
Machine Learning
Improved workflow
Honeywell’s Air Detective is a portable, AI‑powered airborne particle analyzer that uses holographic microscopy and cloud‑based machine learning to identify allergens, spores, and other particles in indoor environments, giving HVAC technicians and environmental testing professionals near real‑time insight into air quality so they can recommend targeted filtration and system upgrades.The handheld device captures and classifies particles on disposable cartridges, replacing the traditional process of sending samples to a lab and waiting hours or days for results, and provides homeowners and building managers with clear images and data to guide decisions about improving indoor air quality. The field‑ready tool enhances service accuracy and supports healthier indoor environments for the hundreds of millions of people affected by airborne allergens worldwide
1
Industrials
Honeywell International
Using AI to speed up and simplify Healthcare Testing
Machine Learning
Improved workflow
Honeywell’s breakthrough Digital Holographic Microscopy technology uses AI to rapidly count and classify microscopic particles or cells, enabling healthcare providers to diagnose infections at the point of care instead of waiting 1–2 days for lab results. The portable device captures holographic images of samples—such as peritoneal dialysis fluid—and uses machine‑learning algorithms to detect infection‑indicating white blood cell patterns, helping patients receive faster, more effective treatment. Because it eliminates the need for complex optics, staining, and lab processing, the system simplifies workflows, reduces delays, and supports broader applications beyond healthcare, including air‑quality analysis and liquid‑sample monitoring.
1
Industrials
KONE Corporation
KONE 24/7 Connected Services - Intelligent PdM
Machine Learning
Increased operational uptime
KONE 24/7 Connected Services is an AI‑driven predictive‑maintenance platform that continuously monitors elevators, escalators, and automatic building doors to detect issues early, reduce breakdowns, and improve people flow in buildings. By connecting equipment to KONE’s cloud and analyzing real‑time performance data, the system identifies emerging faults, prioritizes critical issues, and alerts technicians with precise, actionable insights, enabling faster fixes and more efficient maintenance scheduling. Building owners gain greater safety, transparency, and asset‑lifecycle value through round‑the‑clock monitoring, clear reporting, and fact‑based recommendations, while users benefit from smoother, more reliable journeys. The service has demonstrated strong results, including a more than 70% increase in proactive maintenance activities and a 30% reduction in visible faults during the first two years.
1
Industrials
Mitsubishi Heavy Industries
Gen AI Revamp of On-Call After-Sales Service
Generative AI
Increased productivity
Mitsubishi Heavy Industries revamped its 24/7 U.S. after‑sales phone support by introducing a new gen AI and voice recognition system. The system centralizes customer conversations, automatically routes and transcribes calls, summarizes inquiries with AI, and suggests answers from stored data. This overhaul tripled monthly recorded calls (150 → 500) and improved both speed and consistency of responses while easing operator burden.
1
Industrials
Norfolk Southern
Revolutionizing Railroad Safety with Artificial Intelligence
Machine Learning
Improved safety
From autonomous track inspection to advanced algorithms predicting rail maintenance to powerful cameras and AI models inspecting trains as they pass, Norfolk Southern is building the future of rail safety.
1
Industrials
RTX Corporation
AI makes a leap into the cabin
Computer Vision
Improved efficiency
Galley.ai is an AI‑powered virtual crew assistant developed by Collins Aerospace that uses optical sensors, computer vision, and real‑time inventory intelligence to track items, identify safety hazards, and guide cabin crews through food‑and‑beverage service more efficiently, ultimately speeding up passenger service, reducing crew workload, and improving safety by monitoring bin latches and other galley conditions.
1
Industrials
Schneider Electric
AI-powered home energy management
Machine Learning
Improved user experience
Schneider Electric introduced an AI‑powered feature in their Wiser Home app that automatically optimizes two of the largest household energy loads (water heating and electric vehicle charging) by learning user habits, analyzing weather forecasts, dynamic tariffs, contract limits, and solar production to shift consumption to the most cost‑effective times. This helps homeowners reduce bills, ease the complexity of modern electrified homes, and increase the value of solar installations. Lab simulations showed potential annual savings of €400–500 for solar‑equipped homes and €100–150 for others.
1
Industrials
Schneider Electric
Autonomous Supply Chain
Machine Learning
Cost savings
Schneider Electric's “self-healing” supply chain platform uses adaptive machine learning, big data, and IoT to optimize performance-related parameters in real-time, suggests ways to improve financial and operational performance, and predicts business outcomes.Key benefits:
- 6-day reduction in day-in inventory, resulting in a 10% overall inventory decrease.
- Average 15% yield improvement on specific manufacturing lines.
- Over €100 million generated in value and productivity improved by €30 million.
- Delivery time reduced by 6 days.
1
Industrials
Trane Technologies
Autonomous Building Management with Trane AI Control
Machine Learning
Reduced costs
Trane AI Control, powered by BrainBox AI, is an autonomous, always‑on building management solution that integrates with Tracer® SC+ to continuously predict indoor conditions and optimize HVAC performance in real time. By analyzing live data and adjusting equipment settings automatically, it helps facilities reduce HVAC energy consumption by up to 25% and cut carbon emissions by as much as 40%—all without sacrificing occupant comfort or requiring manual intervention. Designed to work with both new and existing systems, AI Control enhances reliability through predictive insights, accelerates sustainability goals, and delivers straightforward ROI by operating seamlessly in the background as a 24/7 optimization engine.
1
Industrials
Trane Technologies
'ARIA' - conversational building AI agent
Generative AI
Increased operations uptime
ARIA is Trane’s generative‑AI building agent designed to help facilities teams make faster, smarter operational decisions by turning complex building data into clear, actionable insights. It instantly diagnoses HVAC issues, recommends maintenance actions, and provides real‑time and historical data visualizations for remote troubleshooting. With multi‑source integration—including HVAC systems, IoT devices, and weather data—ARIA adapts to each building’s goals, prioritizes tasks based on live conditions, and supports teams globally with 14‑language capability. Accessible via chat or voice on mobile or desktop, ARIA streamlines daily operations while continuously improving through ongoing enhancements.
1
Industrials
Union Pacific
Union Pacific's 'UP Chat' drives efficiency
Generative AI
Improved efficiency
Union Pacific built a secure, Gen AI tool—called UP Chat—to help employees work faster and more efficiently by generating summaries, drafting content, analyzing large datasets, and supporting idea generation, all while keeping company data protected. Unlike public AI tools, which the railroad prohibits for business use due to privacy, security, and accuracy risks, UP Chat runs behind the corporate firewall and follows strict usage policies that require human review, limit outputs to internal use, and prevent automated decision‑making.
1
Industrials
Volvo Group
Advanced Analytics and AI enhance Remote Diagnostics
Machine Learning
Increased operational uptime
Volvo Trucks’ collaboration with SAS focuses on strengthening the company’s Remote Diagnostics service by integrating advanced analytics and AI to process millions of vehicle data records in real time, enabling far more accurate detection, interpretation, and prediction of truck issues. By expanding the range of monitored parts and trouble codes and uncovering hidden patterns in IoT data, the SAS platform helps Volvo identify root causes faster, reduce diagnostic time by about 70%, and cut repair time by roughly 25%, ultimately improving uptime for fleet operators.
1
Industrials
Volvo Group
Boost Safety and Performance with Video Telematics
Computer Vision
Improved safety
Volvo Trucks’ partnered with Lytx to introduce AI‑powered video telematics in Volvo trucks. This system uses machine vision and in‑cab analytics to detect risky driving behaviors, improve safety, and support driver coaching.
The integrated DriveCam system monitors factors such as seatbelt use, following distance, and lane departure, while live‑streaming, cloud connectivity, and optional continuous recording give fleet managers real‑time insight into critical events and operator fatigue or distraction. Designed to reduce installation time and enhance DOT compliance, the solution also helps protect drivers by providing objective evidence in incidents and can be configured for fleets of any size.
1
Industrials
Worley
Accurate and fast data extraction using Machine Learning
Machine Learning
Improved workflows
Worley partnered with Arundo Analytics to expedite data extraction from technical diagrams using DataSeer. DataSeer, is a cloud-based machine‑learning platform that automates the extraction of technical information from industrial engineering diagrams, dramatically reducing manual processing time and improving accuracy. Using deep learning and computer vision, DataSeer can instantly identify instruments, valves, lines, and other components in complex schematics, enabling faster bid preparation, more reliable cost estimation, and the rapid creation of digital twins.
1
Materials
Dow Inc.
Predictive Intelligence Formulation Designer
Machine Learning
Time saving
Formulation Designer is Dow’s predictive‑intelligence tool that helps chemists and product developers rapidly explore, evaluate, and optimize material formulations by combining domain expertise with data‑driven modeling. It streamlines the traditionally iterative, experimental formulation process by enabling users to simulate performance outcomes, compare ingredient combinations, and identify promising formulation pathways before physical testing. Designed to accelerate innovation and reduce development cycles, the tool provides an intuitive interface for navigating complex formulation spaces, improving decision‑making, and supporting the creation of higher‑performing, more sustainable products.
1
Materials
Dow Inc.
Integrated Research Imaging Solution - advancing Coatings
Machine Learning
Improved workflow
Dow’s Integrated Research Imaging Solution is a unified digital platform designed to accelerate coatings innovation by bringing together advanced imaging, materials analysis, and data‑driven insights in a single workflow. It enables researchers to visualize coating structures with greater clarity, correlate micro‑scale features with macro‑scale performance, and make faster, more confident formulation decisions. By integrating imaging tools, analytical models, and collaborative data environments, the solution reduces experimental cycles, enhances understanding of material behavior, and supports the development of higher‑performing, more sustainable coating technologies.
1
Materials
Givaudan SA
Customer Foresight: looking beyond the horizon
Machine Learning
Better product strategy
Customer Foresight is Givaudan’s AI‑enhanced futurescaping platform that blends human expertise, big data and advanced analytics to anticipate long‑term shifts in food and beverage experiences, helping brands move beyond trend‑following into true strategic foresight. By detecting weak signals, mapping drivers of change and generating scenario‑based insights, Givaudan’s futurescapers use the platform’s proprietary digital engine to translate massive, complex datasets into clear patterns and actionable opportunities. The result is a co‑creative process that enables customers to understand emerging consumer expectations, explore future market directions and design next‑generation product experiences with confidence.
1
Materials
Heidelberg Materials
'Heidi' - AI voice agent supporting Order Management
Generative AI
Improved user experience
Heidi is Heidelberg Materials’ AI‑powered voice agent designed to streamline order management by letting customers speak naturally to retrieve orders, confirm project details, and make changes directly in SAP, all without waiting for a human agent . Piloted in Australia, Heidi blends generative AI with controlled business logic to ensure accuracy, handing calls seamlessly to customer service when needed and providing a conversation summary so customers need not repeat themselves. Early feedback highlights its natural voice, flexible dialogue, and ability to bypass queues, with intent‑recognition reliability already above 90% and improving. Built through a cross‑functional global–local collaboration, Heidi is now being strengthened with new features and prepared for thoughtful global scaling, with the underlying logic also enabling easy expansion into chat‑based interfaces.
1
Materials
Holcim
Predicting failures with M-Predict
Machine Learning
Increased operational uptime
M‑PREDICT is Holcim’s AI‑powered predictive‑maintenance tool that uses smart sensors to anticipate equipment failures and maintenance needs, strengthening operational reliability across Plants of Tomorrow sites. The platform analyzes sensor data to forecast issues before they occur, enabling proactive interventions that reduce unplanned downtime and optimize maintenance planning. In plants where M‑PREDICT covers most critical equipment, Holcim reports a 6% reduction in specific maintenance costs, a 40% increase in mean time between failure (MTBF), and a 1.2‑percentage‑point increase in equipment availability.
1
Materials
IFF
Smart dosing Robot transforms fragrance production
Robotics
Improved workflow
IFF’s smart dosing robot, powered by the Colibri automation system, transforms fragrance production by compounding sample batches in minutes and dramatically accelerating the path from idea to finished formula. Installed at the company’s Chin Bee facility in Singapore, the robot handles multiple ingredients simultaneously and operates four times faster than the previous setup, enabling up to 200 sample batches in eight hours—work that once required a full day. This leap in speed, precision, and automation shortens time‑to‑market, boosts capacity, and gives perfumers greater agility to experiment and innovate.
1
Materials
IFF
AI Evolution in Ethanol Production
Machine Learning
Improved workflow
IFF’s AI Evolution initiative is reshaping ethanol production by bringing advanced analytics, machine learning, and holistic plant modeling directly into day‑to‑day operations. Diagnostic, predictive, and prescriptive analytics now help plants detect trends early, prevent issues, and make smarter, data‑driven decisions for higher yields. At the center is the XCELIS® AI platform, which analyzes hundreds of data inputs to optimize fermentation and overall plant efficiency, pushing producers closer to theoretical yield limits. While adoption beyond sequencing technologies is still gradual due to cost and risk concerns, AI‑powered plant models and probabilistic decision tools are steadily proving their value—offering ethanol producers a clearer, faster, and more precise path to operational excellence.
1
Materials
Rio Tinto
Using AI to help conserve 'Ringo Starr' bird hollows
Machine Learning
Time saving
Rio Tinto uses AI to accelerate and strengthen conservation of the endangered palm cockatoo by automating the detection of birds and nesting activity across its vast Weipa bauxite operations. By training a YOLOv5 object‑detection model on more than 10,000 tagged images, the company can now rapidly scan tens of thousands of time‑lapse photos from camera traps placed at potential hollows, identifying cockatoos and competitors far faster than manual review. This automation frees ecologists to focus on understanding breeding behaviour, mapping hollow networks, and improving land‑management decisions. The AI system has significantly reduced data‑processing time and strengthened long‑term protection of critical nesting habitat for one of Australia’s rarest and slowest‑breeding parrots.
1
Materials
Saint-Gobain
Saint-Gobain AI ChatBot
Generative AI
Improved user experience
Saint‑Gobain’s AI Chatbot is an on‑site virtual assistant designed to give users quick, first‑level guidance on the Group’s brands, solutions, and career opportunities, offering both text and voice interaction directly from a small icon available on every page of the website. It helps visitors navigate Saint‑Gobain’s broad ecosystem of resources while clearly avoiding sensitive areas such as customer service, investor relations, or financial data, and it emphasizes privacy by not requiring identification and deleting exchanged data weekly. The tool provides convenient, informational responses while directing users to official contact channels for any specific or detailed inquiries.
1
Materials
Vale
Machine Learning-based Slag Furnace ratio automation
Machine Learning
Improved workflow
Vale applies machine learning to stabilize and automate the Silica/Magnesia (S/M) ratio in its slag furnaces, solving a long‑standing bottleneck in nickel smelting by replacing constant manual adjustments with a real‑time, data‑driven control system. Using neural networks integrated directly with PLCs and XRF laboratory data, the system autonomously optimizes conveyor speeds for different ore blocks, improving S/M compliance from 89% to 95% and eliminating delays of up to 150 minutes that previously reduced furnace stability. This automation boosts productivity—unlocking an estimated 182 additional tons of nickel per year—while reducing operator fatigue, strengthening safety, and cutting emissions by as much as 6,000 tons of CO₂eq annually. The project, which earned PT Vale Indonesia a Gold Achievement at OPEXCON 2025, marks a major step in the company’s broader digital transformation and is now being expanded to other areas such as sulfur and coal management.
1
Utilities
DEWA
DEWA provides services through ChatGPT Apps Directory
Generative AI
Improved user experience
Dubai Electricity and Water Authority (DEWA) has become the first government entity and utility globally to offer its services through the ChatGPT Apps Directory, marking a transformative shift in public service delivery. This initiative aligns with Dubai’s vision to lead in AI and future technologies, enabling customers to access DEWA services—such as billing inquiries, account support, EV charger locations, and Customer Happiness Centre details—via natural conversational interfaces. By integrating with AI platforms like ChatGPT, DEWA is redefining digital interaction, enhancing service agility, and reinforcing Dubai’s position as a global hub for smart, secure, and AI-native innovation
1
Utilities
DEWA
'Rammas' - DEWA’s virtual employee
Generative AI
Improved user experience
DEWA’s AI-powered virtual employee Rammas handled more than 1.6 million customer inquiries in 2025, reflecting the utility’s rapid acceleration in digital transformation and intelligent service delivery. Built on advanced conversational AI and enhanced with ChatGPT since 2023, Rammas now provides real‑time, bilingual support across DEWA’s website, mobile app, service robots, and major voice assistants. Since its launch in 2017, it has processed over 12.7 million inquiries, continuously learning to deliver faster, more accurate responses while recommending relevant services and enabling secure transactions such as EasyPay and multi‑account bill payments.
1
Utilities
DEWA
DEWA deploys Intelligence Data Modelling Software
Machine Learning
Improved workflow
Dubai Electricity and Water Authority (DEWA) has deployed its patented Intelligence Data Modelling Software (IDMS) to dramatically accelerate operational response across its water network. The system uses machine learning to analyse large volumes of SCADA data—such as alarms, notifications, and equipment signals—to help teams detect, diagnose, and resolve issues with greater speed and accuracy. By shifting from traditional monitoring to predictive, data‑driven models, IDMS enhances network reliability, reduces downtime, improves asset safety, and cuts operational costs. The solution also automates SMS alerts to relevant personnel, ensuring rapid action during faults or emergencies.
1
Utilities
E.ON
(Opti)heat - Predicting Heating Demand
Machine Learning
Improved margins
E.ON’s Optiheat is an advanced AI‑driven forecasting system that transforms the efficiency of district heating networks by predicting heat demand up to five days in advance with roughly 90% accuracy. By integrating grid models, customer load profiles, weather forecasts, and historical data, the platform enables dispatchers to optimise production, distribution, and consumption in real time—reducing fuel use, lowering CO₂ emissions, and improving system flexibility. Cities across Europe, including Örebro and Szczecin, are already using Optiheat to cut heat losses, avoid overproduction, and even generate additional electricity revenues.
1
Utilities
EDP Group
Using Robots to improve Wind Turbine Maintenance
Robotics
Increased uptime
EDP uses robots and drones to make wind turbine maintenance faster, safer, and more efficient — reducing downtime, protecting their teams, and ensuring continuous operation, even with adverse weather conditions.
Repairing wind turbine blades—once a risky task requiring technicians to work 80 meters above ground and wait for ideal weather—can now be completed in hours using robotic systems with minimal operational disruption.
At sea, advanced robots act as remote “eyes and hands,” enabling inspections in hard‑to‑reach offshore environments while turbines continue running. These innovations cut downtime by up to 60% and significantly reduce human exposure to hazardous conditions.
1
Utilities
EDP Group
Leveraging AI for Smarter Energy Monitoring
Machine Learning
Improved user experience
EDP’s Smart Energy Monitoring initiative leverages big data and machine learning to improve energy efficiency for buildings in Singapore. Instead of relying on costly, intrusive sub‑metering installations, EDP developed a proprietary AI‑powered Non‑Intrusive Load Monitoring (NILM) system that captures real‑time, per‑equipment energy consumption from a single point at the main switchboard. This approach delivers granular, actionable insights at low cost, enabling faster fault detection, better resource planning, optimized equipment usage, and meaningful operational savings. By transforming raw aggregated data into clear, appliance‑level intelligence, the solution empowers both residential and commercial users to reduce energy waste.
1
Utilities
EDP Group
AI enhances Inspection of Overhead Power Lines
Machine Learning
Improved workflow
EDP’s AI‑powered inspection initiative modernizes how overhead transmission and distribution lines are monitored by replacing slow, repetitive manual analysis with automated intelligence. Using helicopters and drones equipped with laser sensors, thermography, and high‑resolution imaging, Labelec collects detailed aerial data, which EDP then processes through machine‑learning models that automatically classify terrain elements and filter out irrelevant objects like birds or debris. This automation frees specialists from low‑value manual sorting and enables them to focus on higher‑impact tasks.
In a later phase, EDP—working with DefinedCrowd—developed cognitive image‑processing algorithms capable of detecting power line anomalies directly from photographs. The result is faster, more accurate anomaly detection, improved operational efficiency, and a safer, more scalable approach to maintaining critical power‑line infrastructure.
1
Utilities
Entergy
Using AI to improve Power Supply Reliability
Machine Learning
Increased uptime
Entergy deployed an advanced artificial‑intelligence system across its service region to significantly improve grid reliability by predicting equipment failures before they occur. By leveraging data from its smart‑meter (AMI) network—essentially a massive real‑time performance database—the company was able to identify distribution transformers at risk of failing and schedule proactive maintenance rather than responding to unplanned outages.In its first year, the AI‑driven approach prevented 536 outages and avoided more than 48,000 minutes of customer disruption, demonstrating the impact of combining AMI data with machine‑learning analytics. Entergy highlights additional benefits such as improved operational efficiency, better load forecasting, enhanced billing accuracy, and streamlined regulatory reporting.
1
Utilities
Eversource Energy
'SmartInspect' - Streamline Inspection of Transmission Lines
Computer Vision
Improved workflow
SmartInspect is Eversource’s AI-powered inspection platform designed to enhance the monitoring of transmission and distribution lines across its service territory. By leveraging advanced image recognition and cloud-based analytics, SmartInspect automates defect detection—identifying issues like vegetation encroachment, rusted equipment, and broken crossarms simply by scanning assets from a moving vehicle.This approach has significantly improved inspection accuracy and efficiency, saving hundreds of hours compared to traditional methods. Recognized with the Edison Award by the Edison Electric Institute, SmartInspect exemplifies innovation, and reliability in grid operations.
1
Utilities
Iberdrola
'MeteoFlow' - forecast and optimize Renewable Production
Machine Learning
Increased uptime
MeteoFlow is Iberdrola’s advanced, fully in‑house meteorological forecasting system designed to predict renewable energy production with exceptional accuracy across wind, solar, and hydro facilities worldwide. Built and refined over more than 15 years, it integrates modern weather‑forecasting science with machine learning, artificial intelligence, and big‑data processing to deliver hourly forecasts up to 96 hours and daily forecasts up to 10 days for more than 450 plants. By combining global weather data, historical production records, and real‑time operational inputs, MeteoFlow enables smarter energy‑market decisions, optimized maintenance scheduling, and improved profitability for renewable assets. Its capabilities span wind, solar radiation, hydraulic flow, and wave forecasting, and the system has earned international recognition—including selection by IRCAI/UNESCO as one of the world’s top AI‑for‑sustainability projects.
1
Utilities
National Grid
OceanBrain - Monitoring Subsea Cables
Machine Learning
Better risk management
National Grid Partners is tackling the challenge of monitoring the world’s rapidly expanding network of subsea power and telecom cables—critical infrastructure that carries most global internet traffic and enables cross‑border clean‑energy exchange—by developing OceanBrain, a data‑driven risk‑assessment platform. The system fuses cable‑route information, burial depth, seabed conditions, satellite AIS vessel‑tracking data, and machine‑learning models to identify when fishing trawlers are actively dragging nets, a major source of potential cable strikes.By quantifying both real‑time and historical risk, OceanBrain helps operators anticipate vulnerable segments, schedule targeted inspections, and reduce outages that could disrupt renewable‑energy flows. The project is moving toward deployment across multiple National Grid interconnectors, with plans to expand its predictive capabilities using additional environmental datasets.
1
Utilities
National Grid
Streamlining analysis of Legal Documents with Gen AI
Generative AI
Increased productivity
National Grid Partners is adopting Luminance’s legal‑grade AI platform to streamline and modernize how the company manages, reviews, and negotiates contracts across its U.S. and UK operations. Built by University of Cambridge experts, Luminance automates contract analysis, highlights risks, accelerates supplier negotiations, and ensures compliance with evolving regulations—freeing teams from manual document review and enabling faster, more informed decision‑making.A successful 2023 proof‑of‑concept revealed immediate cost‑saving opportunities, prompting National Grid to expand the tool’s use across procurement, treasury, insurance, and electricity distribution. As the platform learns from each interaction, it captures institutional knowledge and tailors insights to each business unit, helping National Grid work more efficiently.
1
Utilities
SSE
Species monitoring with AI
Machine Learning
Improved environmental conservation
SSE Renewables, in partnership with Microsoft, Avanade, and NatureScot, developed an AI‑driven species recognition system that uses cameras and machine‑learning models to automatically detect and count wildlife such as puffins and salmon. Initially piloted on puffin populations along the Scottish coastline, the technology has since expanded to additional species and sites, dramatically improving the accuracy and volume of ecological data compared to traditional manual monitoring. The collaboration earned the Scottish Green Energy Award for Innovation, highlighting how AI can strengthen environmental stewardship while supporting renewable energy development. The partners view this as a foundational step toward broader applications of digital innovation in conservation and sustainability.
1
Utilities
Tenaga Nasional
Machine Learning for 11kV Underground Cables
Machine Learning
Increased uptime
TNB transformed the reliability of its 11kV underground cable network by replacing traditional preventive and condition‑based maintenance (which often produced false positives, missed early failures, and contributed to 70% of SAIDI disruptions) with an AI‑driven predictive maintenance model built on comprehensive historical, operational, and geospatial data.Using a stacked ensemble of CatBoost, XGBoost, Random Forest, and Linear Regression algorithms, refined through advanced feature engineering and dimensionality‑reduction techniques, the model accurately identifies high‑risk cables and enables targeted, cost‑effective interventions without relying on expensive real‑time sensors. Trained on data from 2017–2022 and validated on an independent 2022 dataset, the system achieved 91% accuracy, 81% recall, and a 73% F1‑score, proving its ability to pre‑empt failures and enhance grid resilience.Following a successful pilot in Kuala Lumpur and Selangor with over 90% prediction accuracy, TNB is now rolling out the solution nationwide, marking a major leap toward a fully predictive, data‑driven smart‑grid strategy that strengthens operational efficiency and improves service continuity for customers.
1
Utilities
Tenaga Nasional
'WeKnow' - GenAI Assistant
Generative AI
Improved workflow
TNB’s WeKnow is an internal Gen AI assistant designed to help employees navigate the company’s extensive policies, guidelines, and technical manuals while embodying TNB’s commitment to safe, ethical, and responsible AI adoption. Built around principles such as human‑centricity, accountability, fairness, transparency, and strong data governance, WeKnow enhances operational consistency by providing fast, accurate guidance and helping employees follow proper procedures aligned with TNB’s safety and sustainability standards. The system supports hands‑free voice interaction, generates documentation outputs, and reinforces compliance without replacing human judgment, ultimately improving efficiency and decision‑making.
