Empowering Utility Providers with Computer Vision: Right First Time Installations and Operational Efficiencies

Introduction

As utility providers face increasing pressure to complete their smart meter roll-out programs, computer vision and image recognition can be used to help achieve key operational KPIs such as roll-out rate, quality assurance and cost efficiency targets.  From gas and water distribution to electricity supply, computer vision is transforming how the utility industry operates, bringing precision, automation, and actionable insights to operators that are using these capabilities.

Smart Meter Deployment

The rollout of smart meters is critical for modernizing utility infrastructure, improving energy efficiency, achieving sustainability goals and enabling more dynamic pricing models for customers to adapt their usage according to capacity and cost of supply.

Governments worldwide have set ambitious goals to accelerate the deployment of smart meters. Whilst the EU first mandated its member states to roll out smart meters in 2009, roll-out varies widely by market. For example, Germany lags behind the European average with smart meter adoption at less than 25% of total premises, whilst households in Sweden and Spain have already reached 100 percent deployment.

‍Governments such as the UK have also scaled back their targets as utility providers fail to deploy at the rate anticipated. In the UK, smart meter penetration is now targeted to reach 75% of premises by the end of 2025 vs. the 100% goal originally set out. While there has been a slowdown in some markets, worldwide deployment continues at pace, with smart meter deployment (gas, electricity and water) expected to grow from 1.7 billion in 2023 to 3.4 billion meters by 2033.

Key Challenges:

With providers challenged to meet their roll-out KPIs, computer vision can be used to address operational deployment challenges faced in the field today. Issues include:

• Faulty installations: Right First Time quality installation rates vary significantly by operator. At best case, issues get picked up at the audit stage, triggering a repeat visit to resolve, through to the worst case where undetected safety issues such as loose connections are an actual fire hazard.

• Reliance on contractor staff / agencies:  Reliance on contractors and agency workers are commonplace, with operators often experiencing noticeable differences in Right first time rates between contracting agencies and their own field engineers, combined with skilled workforce shortages.

• Auditing & Compliance: Typically, utility operators are auditing less than 10% of their installations and completed maintenance operations, meaning they only capture a small  proportion of potential defects and / or non-compliant operations, leading to later faults and repair costs, or making payments for installs by contractors that do not meet defined SLA standards.

The Opportunity

Computer vision AI helps with utility industry challenges by offering a proven and innovative approach to enhancing the current process for smart meter installations.

• Real time quality assurance: Automated visual inspections verify that meters are installed correctly and functioning as intended before the technician leaves the premise, leading to a significant reduction in site revisits and rework costs, while ultimately improving the customer experience.

• Improved job completion reporting: Each installation task can be visually documented, creating a clear audit trail for compliance reporting and internal reviews, as well as ensuring that photographic images conform to basic quality standards. Completion reports can also be automatically filled out by using vision AI, so technicians no longer waste time writing down every serial number and asset ID that requires documentation – all it takes is a photo.

• Scale compliance checks: Auditors can now use vision AI to quality check every submitted completion report, leaving the existing audit team to focus on managing the identified issues and exceptions. This can be done without scaling current auditing teams or physically attending the customer premises.  

• Performance management: Increasing the visibility of field technician performance facilitates the proactive training and management of technicians and contractors that fail to meet the required standards. This leads to a reduction in and cost of investigating field technicians that fail to meet performance.

• Improved economics & SLA management: Improve the economics of installing smart meters, through reducing revisit costs and ensuring that contractors are only paid for operations that meet predefined pass criteria.

• Empowering technicians: Computer vision can also be used to guide technicians during the installation or repair process leading to faster resolution, particularly if technicians are working on equipment that is new or they have not been adequately trained on.

‍

Example Installation Checks with Computer Vision

Computer vision can assist in performing many key checks during the smart meter installation process. A few examples for electricity smart installation are outlined below:

• Reference/Label Reading: Using AI to read and verify labels or reference codes on meters to ensure correct identification and configuration.

• Cable Sequence: Ensuring that tails from the cut-out to the meter follows the correct sequence to avoid functionality issues or safety hazards.

• Screw Depth: Verifying that screws are properly tightened to secure components without causing damage.

• Earth Wiring: Confirming the correct placement and connection of earth wires to ensure safety and compliance.

• Isolation Switch – Validating the correct installation of the isolation switch.

• Smart Meter Placement: Checking the position and alignment of the meter to ensure it meets installation guidelines and operational requirements.