Automated Quality Control
StratusGrid and AWS teams are working together to improve the cloud experience, including automating quality control in product manufacturing pipelines or detecting anomalies in the energy/utilities industry. Our Solutions Architecture team built a solution using Amazon Lookout for Vision and created technical training resources.
StratusGrid Services
StratusGrid engineers can deliver projects that utilize the core computer vision capabilities in AWS, and build integrations surrounding it. Integrations that may be useful to customers in this space include:
Notification / alerting mechanism of newly identified anomalies
Extracting and processing individual frames from video
Long-term storage of analysis in databases, for auditing
Using drone fleets for automated data collection
Gathering and resampling images and video for long-term archival
Our Approach to Computer Vision
StratusGrid builds off of cloud services provided by partners such as Amazon Web Services (AWS), and either integrates with them as-is or customizes them to specific customer use cases. We can develop computer vision integrations that utilize object detection and anomaly detection.
Problem Analysis
We start by assessing your end-to-end business model, and then identify opportunities to automate and improve quality controls, using machine learning models designed for computer vision tasks.
Model Training
If model fine-tuning / training is required, identify the data required to train a custom model version, and develop a data quality and validation pipeline. Then, we train a custom model and deploy it for production inferencing.
Integrate
We develop custom integrations, using cloud native services, leveraging computer vision, storage, analytics, and alert management tools, based around industry leading toolchains.
Why Automate Quality Controls With Computer Vision?
Accuracy
Computer vision can accurately detect important real-life events, such as anomalies in manufacturing pipelines, defects in above-ground energy transmission lines, and related equipment.
Automation & Scale
By automating anomaly and object detection tasks with cloud services, you can scale your business operations virtually infinitely. Automated computer vision scales more efficiently than hiring additional dedicated labor for quality control purposes.
Extensibility
Machine learning models built for computer vision are customizable, and can be adapted to your specific use cases.
How We Do It
Containers
Deploy long-running services, using an industry standard, cross-cloud application packaging format.
Serverless
Develop small functions (microservices) to fulfill and scale individual components of the overall computer vision solution.
Amazon Rekognition
Fully managed service that provides APIs for object detection and image classification, and training on custom labels.
Amazon Lookout for Vision
Fully managed computer vision service for anomaly detection, supporting training on custom datasets.
Amazon DynamoDB
Store semi-structured results from computer vision tasks, in a high performance key-value store.
AWS Step Functions
Orchestrate data ingestion, transformation, detection, and results storage.
Industries We Serve
Media & Entertainment
Get next-gen content delivery that gives you a competitive edge. Our team has experience building scalable streaming platforms, expanding them to serve new regions, and ensuring they are always responsive and stable.
Learn About This Industry
Supply Chain Logistics & Advanced Manufacturing
We became an AWS supply chain competency launch partner in 2022. We support digital transformation, including cloud migration, modernization, and stabilization. Our team helps you enhance your digital products at scale.
Learn About This Industry
Healthcare Tech & FinTech
It’s mission-critical to protect personal information for finance and health-based institutions. StratusGrid is SOC2 Type 2 certified and HIPAA compliant, helping you build a secure architecture while ensuring rapid feature development.
Learn About This Industry
With a deep understanding of FinOps, the Stratusgrid team have developed a unique application that provides insightful analysis to help drive cloud cost efficiencies and spot opportunities for architectural improvements. The team’s balanced understanding of SaaS business and technology drivers is a real differentiator."