Modern Manufacturing

Demonstrations of Personalized Smart Manufacturing

John Klinke
Director, Oracle Industry Strategy Group

Guest Author: Surya Kommareddy

Last month, I wrote about the trends and technologies driving hyper-personalized manufacturing. Today, I’d like to illustrate how such a smart factory geared for hyper-personalized manufacturing can be implemented. We recently demonstrated two different concept factories of the future for personalized manufacturing. In these demos, visitors could place an order for a customized luggage tag or a Lego gearbox and experience the end-to-end process of placing an order through to the completion of the product assembly.

Customized Luggage Tag Demo:

Lego Gearbox Demo:

Both demos involved three operations: 1) an initial production stage, 2) movement and quality check of work in progress (WIP) and inventory by an autonomous mobile robot, and 3) a guided final assembly of the customized product at a poka-yoke station. 

The production begins with placement of an order for a customized product from a potential list of choices. Once the order is accepted, a corresponding work order is created in Oracle Supply Chain Management Cloud with the appropriate inventory. The work order is released for production and sent to the first operation. Once the first operation is complete, WIP is transferred to the autonomous robot for AI vision-based material checks and then routed to the poka-yoke machine for final assembly. The final assembly is done under guidance at the poka-yoke station to fulfill and complete the personalized order. Below we will explore some of the functionality demonstrated and how manufacturers can leverage these technologies.

Concepts Demonstrated:

  1. IoT Monitoring: In the luggage tag personalization demo, Oracle IoT Asset Monitoring monitored laser engraver sensor data such as temperature, axis positions, motor speed, and fan speed. Rules were set to notify when the laser temperature went above a limit. Similarly, machine learning models can be trained to identify anomalies based on historical data. Such real-time feedback facilitates rapid responses leading to lower unplanned downtimes and increased asset performance.
  2. IoT Digital Twin: A digital twin of the actual laser engraving operation was built into the IoT Asset Monitoring so one could see digital representation of the engraving toolpath progress and status. In cases where the actual process of manufacturing is not visible, the digital twins can help visualize in real-time what is happening with that asset or component. The digital twin could also be used for cycle time studies and other hypothetical planning tasks without the actual physical asset.
  3. IoT AR Monitoring: An AR application was used to monitor the internal components of the operating engraving machine. This allows anyone who is not familiar with the internal construction of the machine to visualize the components overlaid with operating details. Equipment operators and field service staff can use AR based technologies to visualize, troubleshoot, and resolve problems faster.
  4. OT-IT Integration: The laser engraving machine, the autonomous mobile robot, and poke yoke station were integrated to the Oracle SCM Cloud for transacting work order progression and material movement. Tighter integration between the physical and digital ensures accurate, real-time information flow between systems of planning, execution, and record keeping enabling dynamic readjustment of production. 
  5. Autonomous Robots: Autonomous mobile robots from Omron and Waypoint were used in these demos to move inventory and WIP between operations. Usage of autonomous robots that are capable of programmable movement and decision making helps achieve operational flexibility and agility. In dynamically changing factory floor environment these autonomous robots can work independently with least interruption.
  6. Flexible Routing: The autonomous robot program could be remotely updated to dynamically change the routing to different waypoints based on new production requirements. With more autonomous robots on the factory floor, it is now possible to dynamically push programs and change behavior of the robots, path planning, and routing to implement flexible operations.
  7. Computer Vision-based Detection and Quality Control: The autonomous mobile robots from Omron and Waypoint were rigged with a structure to hold an on-board camera for product detection and identification with the help of Oracle Machine Learning microservices. When issues were detected, a quality exception was automatically created in Oracle SCM Cloud and rerouted accordingly. With AI and high-resolution cameras, it is now possible to perform in-process visual quality checks and make decisions reliably further automating processes and work order execution. Detecting quality issues early on and correcting is critical to delivering batch-of-one products with high quality and on time.
  8. Error-proof Guided Assembly: Since the final assembly of luggage tag is to be performed per the custom order placed, the Mitsubishi Poke Yoke station was used not only to provide the operator with order related information, but also to guide the operator on which inventory to use and how to assemble so as to reduce errors and wastage. With ubiquitous sensors providing feedback and a digital thread connecting workers, assets, and business systems, it is now possible to drive lean, error-proof batch-of-one operations.
  9. Supply Chain Integration: The machines used in the demos transacted and updated their operational status directly in the Oracle SCM Cloud. As more autonomous machinery participates in the manufacturing processes, integrating those assets into the manufacturing operations management and tracking every step of the production is important to monitor progress, dynamically configure routing as necessary, accurately manage inventory, ensure timely delivery, and immediately remediate or respond to any production issues.

Increasingly, the world of manufacturing is racing towards small batch production, mass customization, or a batch of one. Investing in dedicated production lines, automation and multi-tasking machines can increase productivity, quality and delivery, but not necessarily provide the flexibility required to execute hyper-personalized production. Manufacturers can now achieve the vision of hyper-personalization by adopting modular production cells inter-connected with flexible automation, such as autonomous mobile robots, and implementing a digital thread that not only connects the factories to the enterprise but also the suppliers and customers. Manufacturers need a complete end-to-end ecosystem of solutions to truly achieve hyper-personalized manufacturing and Oracle in one of the few vendors that can provide this. For more information, visit Oracle Smart Manufacturing.