Gusto's Journey to Data-Driven Operations using AWS
Summary
-
Challenges Faced by Industrial Customers
- Data is fragmented and lacks context
- 80% of the time is spent gathering, ingesting, and preparing data
- Building a solution requires significant time and resources, and scaling it is costly
- Customers want easy access to their data regardless of its source
-
AWS Approach: The Industrial Data Fabric
- Enables a data-centric approach throughout the data lifecycle
- Provides abilities such as accessibility, interoperability, sensibility, versioning, and scalability
- Utilizes various AWS services and partners to ingest, store, and actualize data
-
Gusto's Journey
- Gusto, a meal kit company, faced challenges in ramping up production due to fluctuating equipment availability
- They aimed to detect downtime quickly, minimize mean time to repair, and maximize time between faults
- Gusto implemented a solution called "Factory OS" using AWS services and partners:
- Ingested data from PLCs using Hyte
- Routed data to AWS IoT Core for cloud connectivity
- Built a fault detection system using AWS IoT Events
- Implemented predictive maintenance using sensors and anomaly detection
- Developed a real-time Scada system using Ignition
- Integrated OEE data into the system for visibility and decision-making
- Created a digital twin and simulation capabilities to test scenarios and optimize operations
-
Next Steps
- Introduced containerization and GitOps for infrastructure resilience
- Developing a "Factory API" to abstract the factory from vendor-specific implementations
- Exploring the use of AI/ML for predictive maintenance and optimization
-
Recommendations
- Start small, deliver value, and iteratively scale the platform
- Leverage existing standards and tools, avoid building everything from scratch
- Bring cross-functional teams together to foster collaboration and learning
- Encourage the tech team to immerse themselves in the factory environment
The key takeaway is how Gusto leveraged AWS services and partners to build a data-driven, integrated system that enabled them to improve operational efficiency, reduce food waste, and increase productivity - all while preparing for future advancements in predictive maintenance and AI/ML-driven optimization.