TCS Machine Vision platform
10 MINS READ
Field assets are the lifeblood of utility companies.
Decarbonization and the race to achieve net-zero targets have placed the spotlight on aging infrastructure and the importance of more intelligent asset inspection and vegetation management.
Secondary assets such as poles, crossarms and insulators, if neglected, increase the risk of failures in transmission and distribution. Vegetation encroachment in the grid corridor also leads to short circuits and outages, if not managed early. In addition, most utilities still rely on manual, subjective interpretation of these anomalies by sporadic analysis of foot patrol reports and images.
Better use of data and automation is essential to prevent brownouts and blackouts of grids. TCS’ Machine Vision platform, especially curated for utilities and other entities, addresses the need to automate and standardize field asset inspection and vegetation management, and ensure more accurate detection of anomalies. By leveraging more effective image analytics, utilities can cut cost, improve asset life and reduce risks related to faulty assets.
The changing energy ecosystem requires utility companies to reimagine their grid operations for improved reliability, safety, cost-efficiency, and resilience.
Climate change-related accidents such as wild-fires, cyclones, and deep-freeze underscore the need for better management of risk-prone assets and corridors. However, as late-adopters of digital technologies, Utilities need to address several challenges. For instance, they rely mostly on subjective human interpretation and seldom have protocols for image capturing to detect anomalies. Indexing and storing unstructured image data is also a challenge which in turn impacts analytics and visualization. Sometimes, image capturing and ingesting mechanisms fail to distinguish between usable and non-conforming data such as pixelated or blurred images. In addition to gaps in image capture, the lack of autonomous flight path guidance and control for drones makes navigating harsh, cluttered environments and extreme weather conditions, more challenging.
TCS’ Machine Vision platform provides an integrated stakeholder engagement for asset inspection and utilities vegetation management. The core AI-ML solution is augmented with a data-driven prioritization process, image protocol adherence, edge analytics and field work process/annotation management for improved upstream operations. It also supports enterprise ecosystem integration with meaningful visualization and initiation of repair and maintenance workflow for downstream processes. Additionally, a conflated database expedites complex analytics and decision-making.
TCS’ Machine Vision platform includes the following key features:
Prioritization module: It has a configurable prioritization module for both asset inspection and vegetation management that enables utilities to zero in on assets and corridors based on defined business rules.
Image capture and edge analytics: It also includes protocols for image capture and automated ingestion of RGB, LiDAR, thermal, and satellite imagery, with auto rejection of non-conforming images.
Integrated map-view and work force management: TCS’ Machine Vision platform comes with role-based access for workorder generation, supervisor allocation, field inspection, and image upload. An in-built map supports visualization and can be accessed through a web browser, tablet or mobile device.
AI-ML models: The patented AI-ML Models in the platform can automatically identify, segregate and label individual assets after detecting damages and anomalies. Numeric values are assigned to assets based on the damage severity for data-driven prioritization of repair/replacement work.
Encroachment detection: The solution automatically identifies vegetation growth patterns and encroachments from 3D point clouds, drones and satellite imagery.
TCS’ Machine Vision platform offers several benefits related to cost and time savings in addition to better risk management.
Increased time savings: Improved decision-making eliminates rework, unnecessary re-inspections and backend analysis, leading to at least 30-50% time savings.
Improved accuracy of anomaly detection: Utilities can improve accuracy of anomaly detection by up to 60% compared to manual efforts.
Increased cost savings: Due to reduced truck-rolls, lesser dependency on skilled resources and improved asset life, utilities can achieve a minimum 50% cost savings across operations in the first year of production deployment.
Enhanced field operations efficiency: In the very first year, utilities can also improve the efficiency of in-field inspections and operations by nearly 70%.
The TCS advantage
In one of their recent blogs, ARC Advisory Group has highlighted TCS Machine Vision platform’s image analytics capabilities for both utility asset and vegetation management.
Robust AI-ML algorithms: TCS’ Machine Vision platform leverages prebuilt, containerized and outcome-oriented algorithms which have been developed and patented by TCS’ research and utilities innovation labs. These algorithms automate anomaly identification and damage detection. Custom-developed algorithms in the platform also enable utilities to accurately monitor and estimate vegetation and forest corridors through satellite images.
Downstream integration: The solution offers in-built APIs which integrate GIS, asset and workforce management, and other enterprise applications.
Meaningful insights from conflated data storage: It also includes a built-in plug-and-play conflated data platform offering in-built edge intelligence and an integrated workflow module.
Flexible hosting: This docker platform (with seamless mobility and web-support) can be hosted on the cloud (recommended) or on-premise. It is auto scalable in real time and supports rapid data ingestion and processing.
Integrated stakeholder play: TCS’ Machine Vision platform is an integrated, end-to-end platform with a front-end configurable pipeline. It supports seamless data ingestion from multiple sources, in different formats and enables multi-stakeholder engagement.