Computer Vision is heralded by business analysts as an important, integral part of industry 4.0. The AI in the manufacturing market is expected to grow to up to $14Bn by 2028, according to a recent survey [link]. This is because Computer Vision is now a mature technology with decades of research and productization and successful implementation in factories around the world. But still, it is a niche technology in many parts of the manufacturing world beyond AOI (Automated Optical Inspection) applications. But the times are changing.
Computer Vision in the last 5 years has solved visual recognition problems considered insurmountable just 10 years ago, the development has been parabolic. Tracking almost any object with any camera at any speed has become a reality even on our mobile devices. And so, the race for Computer Vision for tracking manual operations has begun. With Computer Vision no longer used only for fixed heavily-constrained detections such as visual inspection, it is liberated to track and understand human and machine behavior. That will bring the true data revolution promised in industry 4.0.
Tulip Vision is part of this worldwide shift in the manufacturing paradigm. It collects valuable data about frontline workers with minimal intrusion, which is the cornerstone of process optimization.
Why did we build our own Computer Vision engine?
We decided to build Tulip Vision to benefit our community of builders. It was a difficult decision because many Computer Vision products already existed when we started, and have gotten better since then. But we have a unique approach to Computer Vision in industrial settings that follows our usability philosophy:
We believe that computer vision, just like data applications, belongs to anyone and everyone in the factory at any time, starting right now.
We also believe that Computer Vision is an integral part of the holistic process of manufacturing; it’s not an “add-on” or “plugin”.
By putting Tulip Vision inside the greater Tulip platform offering, with tight integration, we ensure Computer Vision tools are available to add automation and quality into manual processes.
Most existing third-party Computer Vision services can be integrated into Tulip today, but they require external moderation and configuration
Tulip Vision configuration requires no coding or Computer Vision knowledge, and nothing but a browser that can access the Tulip administration portal
How does this philosophy manifest in Tulip Vision?
Easy to get started - Easy to get to production - Easy to replicate to another use case.
Tulip’s Vision product is meant to take you from “sketch” to “production”. It is very easy to get started fast and strong, without requiring any prerequisite knowledge or costly tooling. But at the same time, it’s also very easy to re-do, re-create, re-hash any Vision configuration for new applications. From the configuration to app logic, Vision is strictly no-code and can provide immediate value. Tulip Vision configurations are transferable and shareable between apps and stations with a simple assignment.
Manual operations focus
Vision includes detectors that are aimed at manual operations, like reading barcodes and text, detecting the arrival or departure of parts, bins, or tools, and ensuring correct selections are made by operators. But Vision can also solve highly complex problems such as visual defect detection, with an automatic machine learning process. It can replace or augment human labor on the line to do a visual inspection with an easily maintainable tool.
Manufacturing data comes first
Vision offers ways to automatically run time studies by detecting the beginning and end of manual operations and collecting the right information over time. And accurate data is the only thing that can drive accurate decisions. Vision can collect data where it’s expensive to do so otherwise, for example doing it continuously at scale. Getting 100 stations fitted with Vision may give you 500 constant data points at your operation, and the ultimate visibility into bottlenecks and process anomalies.
Key Takeaways on Tulip Vision
Computer Vision for manual frontline operations is an advancing reality
Tulip Vision can help floor and line managers take hold of their operations by installing simple and effective visual detection tools
Collecting data about your operations is key for truthful KPI monitoring and process optimization
Vision is a versatile tool for data collection that can continuously monitor human and machine operation
Vision can also automate manual processes by detecting and streamlining a complex multi-step operation, reducing the interaction points with work instructions or inputs from the operator
Tulip Vision can be a gateway to introducing Industry 4.0 to your factory with minimal interruption and effort. To get started with Tulip Vision:
Start using Tulip Vision right away by referring to our getting started guide
Use the pre-built library apps that use Tulip Vision for basic example tasks, and modify them to solve the problems in your operations: Vision Unit Tests | Frontline Operations Platform
Run through our Tulip University course that goes in-depth on all Vision topics: Tulip Vision 404
Check back on Tulip Vision frequently to discover more features that will enable additional ways to do data collection and introduce quality and speed to your processes.
See why manufacturers trust Tulip to help drive digital transformation
Learn how manufacturers are using our no-code platform to connect the people, machines, and systems across their operations with a 30-day free trial.