Labelbox > Case Studies > Sharper Shape's Efficient ML Pipeline with Labelbox and Valohai

Sharper Shape's Efficient ML Pipeline with Labelbox and Valohai

Technology Category

Analytics & Modeling - Machine Learning
Platform as a Service (PaaS) - Application Development Platforms

Applicable Industries

Education
Utilities

Use Cases

Computer Vision
Time Sensitive Networking

Services

Data Science Services
Training

About The Customer

Sharper Shape is a technology company that creates safe, efficient transmission and distribution solutions for utilities. They use drones to perform utility inspections and use computer vision models in advanced aerial sensor systems to power the automatic collection and analysis of unmanned aerial inspection data. Their technology is commonly used for the identification of dangerous setups with electric wiring, such as vegetation growing too close, broken insulators, and more, so that utility companies can find and address potential hazards. As a company fueled by AI, Sharper Shape sets itself apart with a strong, established pipeline for developing their ML models.

The Challenge

Sharper Shape, a company that creates technology for safe, efficient transmission and distribution solutions for utilities, was facing challenges in developing their machine learning (ML) models. The company uses computer vision models in advanced aerial sensor systems to power the automatic collection and analysis of unmanned aerial inspection data. A common use case for their technology is the identification of dangerous setups with electric wiring, such as vegetation growing too close, broken insulators, and more. However, training multiple computer vision models required a vast amount of accurately labeled images. Prior to using Labelbox, the Sharper Shape team relied on heavily manual workflows and experimented with open-source labeling tools that did not provide the required amount of configuration needed for their needs. Additionally, each data scientist had spent up to a third of their time on infrastructure and experiment management.

The Solution

Sharper Shape turned to Labelbox to streamline the labeling process, enabling them to use an array of data types, including tiled imagery, and organize their existing data. With Labelbox, the team could connect their raw data into Labelbox via a simple API. Labelbox’s collaboration features also enabled rapid onboarding, training, and throughput for both internal and skilled external labelers to work together in one centralized environment. In a new initiative, the Sharper Shape team is accelerating their labeling process even more with model-assisted labeling, which allows teams to import their model into Labelbox and address edge cases. After their data is fully annotated inside of Labelbox, their data is exported to the Valohai MLOps platform, where the Sharper Shape team runs their machine learning experiments and training pipelines. Valohai enables Sharper Shape to train their models on powerful cloud hardware without DevOps support and to house all their collaborative experiments under a single application. Established ML processes can be fully automated into Valohai pipelines, so models can be trained each time new annotated data is available from Labelbox.

Operational Impact

The implementation of Labelbox and Valohai has significantly improved the efficiency of Sharper Shape's operations. The use of Labelbox has streamlined the labeling process, enabling the team to use a variety of data types and organize their existing data more effectively. The collaboration features of Labelbox have also facilitated rapid onboarding, training, and throughput for both internal and external labelers. On the other hand, Valohai has enabled Sharper Shape to train their models on powerful cloud hardware without the need for DevOps support. It has also allowed all collaborative experiments to be housed under a single application, and established ML processes to be fully automated into Valohai pipelines. This has resulted in a more efficient and cohesive process for developing ML models.

Quantitative Benefit

Reduced time spent on infrastructure and experiment management by data scientists by up to a third
Onboarding time for new team members reduced to a quarter of the previous time
Enabled efficient collaboration and training of both internal and external labelers