- INTRODUCTION
Most of the Artificial Intelligence & Machine Learning (AI & ML) powered solutions are based on the Sense à Think à Act paradigm. The feeds collected from the Cameras installed in the Warehouse often called Computer Vision can be processed using AI and Deep Learning techniques to derive intelligent insights that can be used for immediate action.
- COMPUTER VISION
There are two terms ‘Machine Vision’ and ‘Computer Vision’ that are widely referred to while talking about innovations being introduced in business applications.
Machine Vision is the ability of machines to see and act. This idea is not new and has been the stuff in science fiction for a long time, it is very much a reality now.
Machine vision is an engineering-based system that uses existing technologies to mechanically ‘see’ steps along a production line. It helps manufacturers detect flaws in their products before they are packaged, ensure that goods are correctly labeled.
If we think of machine vision as the body of a system, computer vision is the retina, optic nerve, brain, and central nervous system. A machine vision system uses a camera to view an image whereas computer vision is the algorithms that process and interpret the image. A machine vision system requires computer vision and does not work without a computer and specific software.
Computer vision can process 3D and moving images as well and perform complex operations to detect features within an image, analyze them, and provide rich insights.
The rapid increase in business solutions leveraging computer vision can be attributed to the following drivers:
-
- Advances in Artificial Intelligence (AI)
- Increasing computational speed
- Decreasing the cost of robotics & infrastructure
- Increasing labor cost
- TYPICAL CHALLENGES IN A WAREHOUSE
-
- Lack of visibility into Warehouse operations leading to interleaving Task delays
- Rework in Put Away tasks due to delays in Pickup
- Delay in unloading/shipment due to the unavailability of dock and unloading/loading areas
- Delays in Pick & Pack due to non-availability of items at designated locations
- The manual effort for Carton/Package Dimensions Measurement
- Increase in Labour cost due to manual scanning, verification, and reconciliation of Goods
- Products on loading docks vulnerable to damage by forklifts and bumpy terrain
- HOW COMPUTER VISION AIDED BY AI & DEEP LEARNING CAN HELP?
Warehouse operations typically have tasks that are interleaved. The tasks are assigned to different workers with different roles. Unless a task is completed by one worker, the next task can not be initiated which could lead to delays.
The following section describes some scenarios where Computer Vision can be used.
4.1 Unloading
-
- Tracking Dock Doors and Unloading areas for availability to send alerts for receiving the next Truck. It can also be used to inform the transporters real-time status of Dock Doors for reducing wait time.
- Auto verification of Package Labels and extraction of Label content to reconcile goods unloaded with receipts
- Inspection of received goods for damages
- Perform preliminary checks to ensuring that Put Away locations are not occupied to avoid rework for forklift operators
- Provide visibility into available spaces for storage planning
- Auto capture of Box Dimensions for assessment of space needs for storage
4.2 Put Away
-
- Guide workers in quickly locating the Item and ensuring that it is stored in the specified location
4.3 Pick & Pack
-
- Checking availability of Order Items at Pickup location before Pick Task is assigned
- Barcode Recognition & Item Detection to ensure that the right item and quantity is picked and reconciling against Customer Orders
- Recommendations for Carton selection & Carton space optimization
- Carton Label Verification & Damage Assessment
4.4 Shipping
-
- Monitor Loading area and Dock Door clearance for initiating shipment tasks
- Reconciling shipment items against Customer Order
- Using Carton dimensions for Truck Load Planning
- CONCLUSION
Computer Vision can help enhance efficiency, optimize put away and pick & pack processes, and minimize operational delays in warehouse operations.