Digital Twin In Manufacturing

Unlocking Precision and Efficiency: The Power of the Digital Twin in Manufacturing

A digital twin represents a virtual replica of a physical object, process, or system. In the context of manufacturing, this sophisticated digital model is continuously updated with real-time data from its physical counterpart, enabling unparalleled insights, predictive capabilities, and proactive decision-making. Far beyond static 3D models or simulations, a true digital twin creates a living, breathing virtual ecosystem that mirrors the dynamic behavior and performance of its physical twin on the factory floor or throughout its lifecycle. This article delves into the intricacies of digital twin technology, its core components, multifaceted benefits, strategic applications, and practical implementation steps for engineers, operations managers, and manufacturing professionals striving for the next level of industrial prowess.

Understanding the Digital Twin Concept in Manufacturing Excellence

The term “digital twin” was coined by Dr. Michael Grieves in 2002, initially describing the concept for product lifecycle management (PLM). However, its maturation and widespread application have been catalyzed by the advent of Industry 4.0, which emphasizes connectivity, automation, and real-time data exchange. For manufacturing, a digital twin is not a single piece of software but an integrated framework that offers a holistic view of operations, bridging the gap between the physical and digital worlds.

Defining the Digital Twin: A Dynamic Virtual Replica

At its core, a digital twin is a virtual construct designed to precisely mirror a physical entity. This physical entity can be anything from a single sensor, a machine component, an entire production line, a finished product, or even an entire factory ecosystem. The ‘twin’ aspect signifies a persistent, dynamic connection: data flows continuously from the physical asset to its digital counterpart, allowing the virtual model to accurately reflect the real-time status, performance, and behavior of the physical entity. This bidirectional data flow is critical, as the insights gained from the digital twin can then be used to inform, optimize, and even control the physical asset.

Key characteristics that differentiate a digital twin from traditional simulation or CAD models include:

• Real-time Data Integration: It continuously ingests data from IoT sensors, PLCs, ERP, MES, and other systems.
• Bi-directional Communication: Information flows from the physical to the digital, and actionable insights or commands can flow back.
• Lifecycle Representation: It evolves with the physical asset, from design and commissioning through operation, maintenance, and eventual decommissioning.
• Predictive and Prescriptive Capabilities: Leveraging advanced analytics and AI, it can forecast future states and recommend optimal actions.

Evolution and Significance in the Industry 4.0 Landscape

In the context of Industry 4.0, digital twins facilitate:

• Smart Factories: By creating virtual models of entire production facilities, enabling real-time monitoring and dynamic optimization of resource allocation, energy consumption, and workflow.
• Cyber-Physical Systems: They are the core of these systems, where physical components are embedded with computing and networking capabilities, allowing them to interact and collaborate seamlessly.
• Data-Driven Decision Making: Providing a consolidated, real-time view of complex operations, empowering stakeholders with actionable insights.

This integration marks a significant leap from isolated systems to a cohesive, intelligent manufacturing ecosystem, where every component can be monitored, analyzed, and optimized virtually before physical intervention.

Core Components and Enabling Technologies for Digital Twin Implementation

The successful deployment of a digital twin in manufacturing relies on the synergistic integration of several advanced technologies and components. These elements work in concert to create a robust, dynamic, and intelligent virtual representation of physical assets and processes.

Physical Assets and Sensor Networks

The starting point for any digital twin is the physical asset itself – a machine, robot, production line, or even a complete facility. To gather the necessary data, these physical entities must be instrumented with a comprehensive network of sensors and actuators. These include:

• IoT Sensors: Monitoring temperature, pressure, vibration, current, voltage, acoustic emissions, and other critical parameters.
• Vision Systems: Cameras and scanners for quality inspection, object recognition, and positional tracking.
• PLCs (Programmable Logic Controllers): Collecting operational data from machine controls.
• RFID/GPS: For tracking assets, materials, and products within the facility and supply chain.
• Edge Computing Devices: Processing data locally before sending it to the cloud, reducing latency and bandwidth requirements.

The quality and volume of data collected directly impact the accuracy and utility of the digital twin.

Virtual Model and Simulation Platforms

The digital twin’s virtual model is more than just a 3D CAD drawing; it’s a dynamic, multi-domain representation incorporating physics-based models, behavioral algorithms, and historical data. Key platforms and tools include:

• CAD (Computer-Aided Design) & CAE (Computer-Aided Engineering): For creating detailed geometric and functional designs, and for simulating stress, fluid dynamics, and thermal properties.
• PLM (Product Lifecycle Management) Systems: Managing product data and processes across its entire lifecycle, from concept to retirement.
• Simulation Software: Tools for discrete event simulation, agent-based modeling, and system dynamics to model complex manufacturing processes, logistics, and resource allocation.
• Physics-Based Models: Incorporating mathematical equations that describe how components behave under various conditions.

Data Integration, Big Data, and Advanced Analytics

The lifeblood of a digital twin is data. Integrating data from diverse sources and applying advanced analytical techniques are crucial:

• Data Integration Platforms: Solutions that connect IoT devices, MES (Manufacturing Execution Systems), ERP (Enterprise Resource Planning) systems, SCADA (Supervisory Control and Data Acquisition), and other enterprise applications.
• Big Data Technologies: Storing, processing, and managing massive volumes of real-time and historical data generated by manufacturing operations.
• AI/ML Algorithms:
  • Predictive Analytics: Forecasting equipment failures, demand fluctuations, or quality deviations.
  • Prescriptive Analytics: Recommending optimal actions to prevent failures, improve efficiency, or mitigate risks.
  • Diagnostic Analytics: Identifying the root causes of problems.

Communication and Connectivity Infrastructure

Seamless and secure communication channels are essential for the continuous flow of data:

• Industrial Communication Protocols: Standards like OPC UA, MQTT, and PROFINET facilitate interoperability between devices and systems.
• Cloud and Edge Computing: Cloud platforms provide scalable infrastructure for data storage, processing, and advanced analytics, while edge computing enables real-time decision-making closer to the data source.
• 5G and Low-Latency Networks: Enabling faster, more reliable data transmission for critical real-time applications.

User Interface, Visualization, and Control

To make the complex data and insights accessible and actionable, robust user interfaces are required:

• Dashboards and HMIs: Providing real-time visualization of operational status, KPIs, and alerts.
• Augmented Reality (AR) & Virtual Reality (VR): Offering immersive experiences for training, remote assistance, and interactive visualization of the digital twin.
• Digital Thread: A seamless, integrated view of all data related to an asset throughout its lifecycle, accessible through these interfaces.

These integrated components form the robust technological backbone that empowers manufacturers to harness the full potential of digital twins.

Transformative Benefits of Digital Twins in Manufacturing Operations

The strategic deployment of a digital twin in manufacturing yields a multitude of profound benefits that span across the entire product lifecycle, from initial design to end-of-life services. These advantages translate directly into enhanced operational efficiency, reduced costs, improved product quality, and a significant boost in competitive positioning.

Enhanced Product Design and Prototyping

Digital twins revolutionize the early stages of product development by enabling extensive virtual testing and refinement before any physical prototypes are built. This leads to:

• Reduced Prototyping Costs and Time: Designers can rapidly iterate and test hundreds of virtual designs without incurring material or labor costs, drastically shortening time-to-market.
• Optimized Performance: Simulations allow engineers to predict how a product will perform under various real-world conditions, identifying flaws and optimizing designs for durability, efficiency, and functionality.
• Design for Manufacturability (DFM) and Assembly (DFA): Virtual models can highlight potential manufacturing challenges or assembly complexities early on, allowing design adjustments that streamline production processes.

Optimized Production Processes and Throughput

On the factory floor, digital twins provide unprecedented visibility and control, leading to significant process improvements:

• Real-time Monitoring and Control: Operators gain a live, comprehensive view of production lines, machines, and material flow, enabling immediate identification and resolution of bottlenecks or deviations.
• Process Simulation and Optimization: Manufacturers can simulate changes to production parameters, robot paths, or line layouts in the digital twin environment to predict their impact and identify optimal configurations without disrupting live operations.
• Reduced Cycle Times: By identifying and eliminating inefficiencies, digital twins help streamline workflows, leading to faster production cycles and increased throughput.
• Energy Management: Monitoring and simulating energy consumption across the factory allows for optimization strategies that reduce utility costs and environmental footprint.

Improved Asset Performance and Predictive Maintenance

Perhaps one of the most compelling benefits of digital twins is their ability to transform asset management and maintenance strategies:

• Predictive Maintenance: By continuously analyzing real-time and historical data from physical assets, the digital twin can predict potential equipment failures long before they occur, triggering maintenance interventions at optimal times. This shifts from reactive or time-based maintenance to condition-based maintenance.
• Reduced Downtime: Proactive maintenance prevents unexpected breakdowns, significantly minimizing costly unplanned downtime and ensuring higher asset availability.
• Extended Asset Lifespan: Optimized maintenance schedules and operational parameters, informed by the digital twin, help extend the operational life of machinery and equipment.
• Remote Diagnostics and Troubleshooting: Technicians can remotely diagnose issues using the digital twin, reducing the need for on-site visits and speeding up resolution times.

Supply Chain Optimization and Enhanced Visibility

Beyond the factory gates, digital twins can extend their influence to the broader supply chain ecosystem:

• Real-time Tracking and Traceability: Creating digital twins of products or batches allows for real-time tracking of their location and condition throughout the supply chain, enhancing transparency and accountability.
• Demand Forecasting and Inventory Management: By integrating market data and production capabilities, digital twins can improve the accuracy of demand forecasts, leading to optimized inventory levels and reduced waste.
• Logistics Simulation: Simulating different logistics scenarios can help identify the most efficient routes, modes of transport, and warehousing strategies, mitigating risks and improving delivery times.

Quality Control and Defect Reduction

Maintaining high-quality standards is critical in manufacturing, and digital twins provide powerful tools for quality assurance:

• Early Defect Detection: Real-time monitoring of production processes allows for the immediate identification of deviations that could lead to defects, enabling prompt corrective actions.
• Root Cause Analysis: When defects do occur, the digital twin can provide a comprehensive historical record of operational parameters, helping engineers quickly pinpoint the root cause and prevent recurrence.
• Process Deviation Monitoring: The twin can flag when a process drifts out of specified tolerances, preventing a batch of non-conforming products from being produced.

Collectively, these benefits underscore how digital twins are not just an incremental improvement but a fundamental paradigm shift, offering manufacturers a competitive edge through unparalleled insight and control over their entire operations.

Strategic Applications of the Digital Twin Across the Manufacturing Lifecycle

The versatility of the digital twin in manufacturing allows its strategic application at every stage of a product’s lifecycle, from its initial conceptualization to its eventual end-of-life. This comprehensive coverage ensures continuous optimization and value generation throughout the entire industrial process.

Design and Engineering Phase: Virtual Commissioning and Validation

In the earliest stages, digital twins are invaluable for validating designs and ensuring manufacturability:

• Virtual Prototyping and Testing: Engineers can create highly detailed digital twins of components and entire systems to simulate their performance under various conditions, stress tests, and environmental factors. This allows for rapid iteration and optimization of designs without the need for expensive physical prototypes.
• Design for Manufacturability (DFM) & Assembly (DFA): By simulating the manufacturing and assembly processes, the digital twin can identify potential bottlenecks, tooling clashes, or ergonomic issues early on, informing design modifications that reduce production costs and time.
• Virtual Commissioning: Before any physical equipment is installed, a digital twin of a new production line or robotic cell can be commissioned virtually. This allows for the testing of control logic, robot programming, and system integration in a risk-free environment, significantly reducing physical commissioning time and associated costs.
• Material Science Simulation: Digital twins can incorporate advanced material models to predict how different materials will behave under specific manufacturing processes (e.g., molding, welding, machining), optimizing material selection and process parameters.

Production and Operations Phase: Real-time Optimization and Control

During active production, digital twins become the central nervous system of the smart factory, providing real-time intelligence and control:

• Real-time Performance Monitoring: A digital twin of a machine, production line, or entire plant constantly aggregates data from sensors, PLCs, and MES to provide a live, accurate view of operational status, throughput, OEE (Overall Equipment Effectiveness), and energy consumption.
• Predictive Quality: By monitoring process parameters in real-time and comparing them against ideal conditions, the digital twin can predict potential quality deviations or defects before they occur, allowing for immediate corrective action.
• Dynamic Scheduling and Resource Allocation: The twin can simulate different production schedules or reallocate resources (e.g., robots, AGVs) in response to unexpected events (machine breakdowns, material shortages) to maintain optimal flow and minimize disruption.
• Process Parameter Optimization: AI-driven analytics within the digital twin can continuously learn from operational data to suggest optimal machine settings, temperatures, speeds, or pressures to maximize yield and efficiency.
• Safety Optimization: By simulating hazardous scenarios or monitoring human-robot interactions, digital twins can identify potential safety risks and inform preventative measures or safety protocol adjustments.

Service and End-of-Life Phase: Predictive Maintenance and Circular Economy

The utility of a digital twin extends well beyond the point of sale, enhancing post-production services and supporting sustainability goals:

• Remote Diagnostics and Prescriptive Maintenance: For products in the field, a digital twin can monitor their performance, predict failures, and provide remote diagnostics. This enables service technicians to arrive on-site with the correct parts and tools, or even resolve issues remotely, significantly reducing service costs and customer downtime.
• Warranty Management and Field Service Optimization: Detailed operational history captured by the digital twin provides objective data for warranty claims, while predictive insights optimize service routes and scheduling for field technicians.
• Product Performance Improvement: By analyzing real-world usage data from digital twins of deployed products, manufacturers gain invaluable insights for future product design improvements and feature enhancements.
• End-of-Life and Circular Economy Initiatives: A digital twin can contain information about a product’s material composition, assembly methods, and wear patterns, facilitating more efficient disassembly, recycling, or remanufacturing processes, thereby supporting circular economy principles.

Through these applications, digital twins ensure that every stage of the manufacturing lifecycle is infused with intelligence, efficiency, and foresight, driving continuous improvement and sustainable operations.

Implementing a Digital Twin Strategy: Practical Steps for Manufacturers

Adopting a digital twin in manufacturing is a strategic undertaking that requires careful planning, a phased approach, and cross-functional collaboration. For engineers and operations managers, navigating this implementation effectively can unlock significant competitive advantages. Here are practical steps to guide the process:

Phase 1: Define Scope, Objectives, and Use Cases

The most critical initial step is to clearly articulate why a digital twin is needed and what specific problems it will solve.

1. Identify Pain Points: Pinpoint critical challenges such as frequent machine breakdowns, high defect rates, energy waste, or slow product development.
2. Define Measurable Objectives (KPIs): Set clear, quantifiable goals (e.g., reduce downtime by 20%, improve OEE by 15%, shorten product design cycles by 30%).
3. Start Small with High-Impact Use Cases: Begin with a pilot project on a critical asset, a specific production line, or a particular manufacturing process that offers a clear, demonstrable ROI. This allows for learning and refinement before large-scale deployment.
4. Secure Executive Buy-in: Present a compelling business case to leadership, highlighting potential ROI and strategic alignment.

Phase 2: Data Acquisition and Infrastructure Development

A robust data foundation is the bedrock of any effective digital twin.

1. Sensor Deployment Strategy: Determine which physical assets need to be instrumented, what parameters to measure, and select appropriate IoT sensors, gateways, and edge computing devices. Prioritize data quality and reliability.
2. Connectivity Infrastructure: Establish a secure and reliable network (wired, wireless, 5G) for data transmission from physical assets to the digital twin platform.
3. Data Integration Strategy: Plan how to integrate data from diverse sources (OT systems like PLCs, SCADA; IT systems like MES, ERP, CMMS, PLM) into a unified data lake or platform. Standardized protocols (e.g., OPC UA, MQTT) are crucial here.
4. Cloud vs. On-Premise Decision: Evaluate whether to host the digital twin platform and data analytics on a public/private cloud, on-premise, or a hybrid model, considering data security, scalability, and latency requirements.

Phase 3: Virtual Model Development and Calibration

Building the virtual representation that accurately mirrors the physical asset.

1. Model Construction: Develop the virtual model using CAD/CAE tools, incorporating geometric data, physics-based simulations, material properties, and operational logic.
2. Data Mapping and Semantic Modeling: Link real-world sensor data to specific parameters within the virtual model, ensuring a clear understanding of what each data point represents.
3. Model Calibration and Validation: Continuously feed live data from the physical twin into the virtual model and compare its simulated behavior against actual performance. Iterate and refine the model to ensure high fidelity and accuracy.

Phase 4: Analytics, AI/ML Integration, and Actionable Insights

Transforming raw data into intelligence that drives decisions.

1. Develop Analytics Models: Implement AI/ML algorithms for predictive maintenance, anomaly detection, process optimization, and quality forecasting.
2. Dashboard and Visualization Development: Create intuitive dashboards, HMIs, or AR/VR interfaces that present key insights, KPIs, alerts, and recommendations to operators and managers in an easily digestible format.
3. Integration with Enterprise Systems: Ensure seamless integration with existing ERP, MES, and other operational systems to enable closed-loop optimization and automated responses.

Phase 5: Deployment, Monitoring, and Iteration

Bringing the digital twin to life and ensuring its long-term value.

1. Pilot Deployment: Roll out the digital twin in the initially defined, limited scope. Gather feedback, identify bottlenecks, and refine the system.
2. Continuous Monitoring and Performance Tracking: Establish a robust monitoring framework to track the digital twin’s performance against its objectives and KPIs.
3. Iterative Improvement and Scalability: Based on learnings from the pilot, iterate on the digital twin’s capabilities and plan for phased expansion to other assets, lines, or even entire facilities. Maintain flexibility for future technology integration.

Key Considerations for Successful Implementation:

• Data Security and Privacy: Implement stringent cybersecurity measures to protect sensitive operational data and intellectual property.
• Interoperability and Standards: Advocate for and utilize open standards to ensure seamless communication between disparate systems and avoid vendor lock-in.
• Talent Development: Invest in training employees in data science, AI/ML, IoT, and digital twin platforms to build in-house expertise.
• Change Management: Prepare the workforce for new ways of working, emphasizing the benefits and providing adequate support.

By following these structured steps, manufacturers can strategically implement digital twin technology, transforming their operations into intelligent, efficient, and resilient ecosystems.

Real-World Impact: Digital Twin Examples in Automotive and Industrial Manufacturing

The theoretical benefits of the digital twin in manufacturing are compelling, but their true power is best illustrated through real-world applications. From the complex assembly lines of the automotive industry to the intricate processes of industrial machinery, digital twins are delivering tangible value.

Mitsubishi Manufacturing’s Approach to Digital Twins (Illustrative Scenario)

As a leader in industrial and automotive production, Mitsubishi Manufacturing exemplifies the potential of digital twins. Consider a hypothetical scenario:

At a Mitsubishi engine assembly plant, critical machining centers are equipped with an array of IoT sensors monitoring vibration, temperature, cutting force, and energy consumption. This real-time data feeds into a digital twin of each machine, which also incorporates its CAD models, historical maintenance records, and material properties. Advanced AI algorithms within the digital twin analyze this influx of data to predict tool wear with high accuracy, often days in advance. This allows maintenance teams to schedule proactive tool changes during planned downtimes, rather than reacting to unexpected failures. The digital twin further simulates different machining parameters—feed rates, speeds, and depths of cut—to identify optimal settings that minimize energy usage while maintaining stringent quality specifications for engine blocks. This not only reduces operational costs but also improves the overall equipment effectiveness (OEE) and extends the lifespan of expensive machinery, demonstrating the core value proposition of a digital twin in a highly competitive automotive manufacturing environment.

Beyond individual machines, Mitsubishi could extend this to entire production lines, simulating material flow, robotic movements, and human-robot collaboration to virtually optimize throughput and identify potential bottlenecks before they impact physical operations.

Aerospace Industry: Predictive Maintenance for Jet Engines

A prominent application is found in the aerospace sector, where engine manufacturers develop digital twins for each jet engine produced. These twins continuously receive data from sensors embedded in the engines during flight – monitoring temperature, pressure, rotational speed, and vibration. The digital twin then analyzes this data against historical performance models to:

• Predict component fatigue and wear.
• Anticipate potential failures and recommend pre-emptive maintenance.
• Optimize flight routes and operational parameters to extend engine life and improve fuel efficiency.

This approach dramatically reduces unplanned maintenance events, enhances safety, and significantly lowers operational costs for airlines.

Automotive Production Lines: Virtual Commissioning and Throughput Optimization

Major automotive manufacturers utilize digital twins for virtual commissioning of new assembly lines or modifications to existing ones. Before any physical equipment is installed, a highly accurate digital twin of the entire line, including robots, conveyors, and human workstations, is created. Engineers and operations managers can then:

• Test and validate robotic programming and sequences.
• Simulate different production scenarios and material flow.
• Identify and resolve bottlenecks in a virtual environment, minimizing costly physical rework and accelerating ramp-up times for new vehicle models.
• Optimize cycle times and balance workloads across stations.

Process Manufacturing: Chemical Plants and Pharmaceutical Production

In industries like chemicals and pharmaceuticals, digital twins are used to model complex chemical reactions, fluid dynamics, and process parameters. This enables:

• Process Optimization: Simulating different operating conditions to find optimal temperatures, pressures, and flow rates for maximizing yield and minimizing waste.
• Safety Enhancements: Predicting potential runaway reactions or equipment malfunctions, allowing operators to implement preventative measures.
• Quality Assurance: Ensuring consistency in batch production by constantly monitoring and adjusting parameters to maintain product specifications.
• Energy Efficiency: Identifying opportunities to reduce energy consumption within energy-intensive processes.

These examples underscore that digital twins are not confined to a single industry but offer universal applicability for any organization seeking to enhance its understanding, control, and optimization of physical assets and processes.

Challenges and Future Trends in Digital Twin Technology

While the digital twin in manufacturing offers transformative potential, its implementation is not without challenges. Understanding these hurdles and anticipating future trends is crucial for manufacturers to effectively strategize their adoption and evolution.

Current Challenges in Digital Twin Adoption

• Data Volume, Velocity, and Quality: Manufacturing environments generate enormous amounts of data at high speeds. Managing, storing, processing, and ensuring the quality and integrity of this data can be a significant undertaking. Inaccurate or incomplete data can lead to flawed insights from the digital twin.
• Interoperability and Standardization: A lack of universal standards for data exchange, communication protocols, and modeling formats can hinder the integration of disparate systems and technologies from different vendors. This often leads to vendor lock-in or complex, bespoke integration efforts.
• Initial Investment and ROI Justification: The upfront costs associated with sensor deployment, software licenses, platform development, and talent acquisition can be substantial. Clearly demonstrating a measurable return on investment, especially in the early stages, is vital for securing continued stakeholder support.
• Cybersecurity Risks: Connecting physical assets to digital networks creates new vulnerabilities. Protecting the digital twin’s data from unauthorized access, manipulation, or cyber-attacks is paramount to prevent operational disruptions, intellectual property theft, or safety hazards.
• Talent Gap: There is a shortage of professionals with the interdisciplinary skills required for digital twin development and management, including expertise in IoT, AI/ML, data science, cloud computing, and domain-specific manufacturing knowledge.
• Complexity of Modeling: Creating highly accurate, physics-based models for complex machinery or entire factories is technically challenging and requires deep domain expertise.

Emerging Trends and Future Directions

Despite the challenges, the evolution of digital twin technology is accelerating, driven by advancements in complementary fields:

• Hyper-realism and Fidelity: Future digital twins will achieve even greater levels of detail and accuracy, mirroring physical assets with unprecedented precision, enabling more nuanced predictive capabilities and finer control.
• Autonomous Digital Twins: Integrating more sophisticated AI and machine learning will lead to digital twins that can autonomously identify issues, propose solutions, and even execute corrective actions on the physical twin without human intervention.
• Metaverse and Immersive Experiences: The convergence of digital twins with augmented reality (AR) and virtual reality (VR) and the emerging industrial metaverse will create highly immersive environments for factory planning, remote maintenance, collaborative design, and training. Operators and engineers could ‘walk through’ a virtual factory, interact with digital twins, and diagnose issues in real-time from anywhere.
• AI-Driven Model Generation and Self-Learning: AI will increasingly automate the creation and calibration of digital twin models, reducing manual effort and continuously improving model accuracy based on new data and operational feedback.
• Digital Twins for Sustainability and Circular Economy: As environmental concerns grow, digital twins will play a larger role in optimizing resource consumption, predicting waste generation, enabling remanufacturing, and managing the end-of-life cycle of products for a more circular economy.
• Interconnected Ecosystems of Twins: The future will see not just individual digital twins, but entire “systems of digital twins” interacting across supply chains, value networks, and even different companies, enabling unprecedented levels of collaboration and optimization.

The journey towards fully leveraging digital twin technology is continuous. By addressing current challenges and strategically embracing these emerging trends, manufacturers can ensure they remain at the forefront of industrial innovation, continually enhancing efficiency, resilience, and sustainability across their operations.