Continuous Improvement with Collaborative AI
<p>Increasing efficiency & safety by using drones for virtual inspections and AI for analysis of the power line maintenance</p>
Drones in the air!
German power supplier needs to go through a long process of checking the power line and maintaining it. Most of the work was done manually with lots of effort. With new digital solutions, the supplier aims to increase efficiency and safety of the process. Drones are used to take photos of the power poles and AI-powered inspection tools analyses the photos.
<h5>Watching with the naked eye</h5><h2>Lengthy evaluation process</h2><p>The organisation is responsible for supplying a certain part of German households with electricity. Previously the inspection of the power line, on a length of about 700 000 km, was done manually by climbing the poles or from a helicopter, simply by subjective evaluation and entering data into the system. Identifying the defect patterns was low and therefore the organisation did not possess with reliable predictions.</p>
<h5>New Solution</h5><h2>Virtual analysis with AI</h2><p>The organization sought process enhancement and chose aerial drone footage coupled with AI-driven virtual image analysis through Microsoft Azure. No need to climb the poles! Technicians fly the drones and acquire pictures of key problematic points. Then the AI comes into place to analyze the pictures, sort and evaluate them.</p>
Efficient Azure Image Workflow
<p>The images taken by the drones are uploaded into Sharepoint and transferred into <strong>Azure Data Lake Gen2</strong>. <strong>Azure Logic Apps</strong> assigns the pictures to each power line and pole. Subsequently, the images undergo refinement via <strong>Azure Functions</strong> and are imported into Cloud-based SaaS Grid Vision® through <strong>Azure Event Hubs</strong> as instances.</p>
Smart! Problems Detected by AI
<p>Various Azure services are involved, with <strong>Azure Kubernetes Service, Azure Database</strong> for <strong>PostgreSQL</strong>, and <strong>Cosmos DB</strong> playing significant roles. Next, AI examines the images, <strong>identifying potential problems</strong> like component issues or power pole damage. An expert user subsequently validates the AI's conclusions before <strong>exporting the results to the Cloud</strong>, where they are transformed into a checklist.</p>
Microsoft AI case studies | No2 | Driving Efficiency
<p>The gathered data enhances the inspections of distribution system operators, enabling the transition from periodic to <strong>predictive maintenance </strong>through identifying defect patterns and making <strong>predictions more accurate</strong>. Grid Vision®, a widely used utility software, employs AI models supported by expert user input to continually train and improve the AI through a global feedback loop, termed collaborative AI.</p>
AI Content Coach
Integrating ChatGPT into Azure OpenAI Service paved the way for more innovations such as AI Content Coach. The interactive assistant trained on the DAM text content facilitates ideation and content creation aligned with a customer's brand voice, advancing the vision of comprehensive content operations and scalable omnichannel personalization.
Proven competence: Awards & Certifications
Discover Technologies and Services
- Technologies
- Services