i-5O Announces Strategic Partnership with Cohere to Enhance AI-Powered Manufacturing Solutions Read More

Background Image

Digital kanban systems :

Why computer vision is the best option

BY KHIZER HAYAT

4 MINUTE READ

Kanban systems are an important inventory control mechanism in manufacturing. Kanban is a Japanese word that directly translates to "visual card", so a kanban system simply means to use visual cues to prompt the action needed to keep a process flowing. These systems are used in manufacturing to indicate when inventory is running low at a particular production line or station so that it can be replenished to prevent any production downtime at the station.

Traditional kanban systems rely on material visual cues such as cards or flags that are raised when inventory is low. Kanban systems differ from poka-yoke systems in that poka-yoke systems are used to prevent errors in production while kanban systems are for avoiding production halts due to a shortage of parts. However, the more modern kanban systems involve the use of electronics to produce digital signals indicating low inventory.

While the modern electronic kanban systems are an upgrade on the traditional ones, they still possess implementation challenges. A much better approach is to use computer vision for tracking inventory for kanban systems as this method costs far less than electronic systems while producing equivalent or better results.

Major challenges with traditional kanban systems

Traditionally, the Kanban is a physical signal in the form of a tag or label, an empty bin or rack position, a marked space on the floor or other physical signal. In the case of a tag, Kanbans can be attached to finished goods, for example. When an item is sold, the Kanban tag is removed and sent back to the final assembly area where it acts as authorization to make another item to replace the one just sold.
Physical Kanban is hampered by any delay in signal recognition and communication. For example, Kanbans often sit until a floor worker who circulates through the plant picks them up and delivers them. Depending on the availability of the worker there can be significant delays which can result in production downtime leading to loss of revenue for the organization.

How modern electronic kanban systems are implemented today

The trend today is for ERP (Enterprise Resource Planning) systems to incorporate an electronic version of Kanban, or eKanban. With the new technology, a replenishment signal can be triggered in real time by an inventory transaction, such as an item sold or issued to production, and communicated directly to the production area, warehouse or supplier, avoiding the aforementioned delays. Devices like sensors are set up on inventory racks to monitor inventory levels and when the level falls below a certain threshold the sensors automatically send a signal to the ERP system for replenishment.

Modern electronic kanban systems are expensive and lack reliability

The major challenge with the implementation of modern electronic kanban systems is that they are expensive to implement. The production plant infrastructure needs to be changed to accommodate sensors that monitor inventory levels and in many cases it becomes impossible to accommodate all the changes. Furthermore, the addition of these sensors often results in production downtime since it can take days to weeks to install them across an entire production plant. A cheaper option is to use barcode readers to scan inventory items when they are moved to a production line however, this method relies on humans doing the scanning which can result in errors in cases where workers forget to barcode an item.

Computer vision is the best electronic kanban system

With the recent developments in the field of computer vision, they have become a far superior option for tracking inventory in manufacturing. Computer vision kanban systems are non-invasive and require a lot less infrastructure adjustments compared to other electronic kanban systems while producing more accurate results. Cameras can be installed on the production floor with minimal to no production downtime and these cameras can then be used to monitor inventory all across the floor.

How i-5O is using computer vision for kanban systems

At i-5O we’ve successfully implemented a computer vision based kanban system at multiple production plants resulting in significant revenue increases for our clients. Our system tracks inventory levels on production lines and as soon as the inventory falls below a certain threshold the system triggers an alert via text, email or on a monitor installed on the production line. Production workers and managers can then take action immediately to replenish the inventory levels. Since the system is self-contained it doesn’t have to channel the alerts to an ERP system and can operate in a fully autonomous fashion. With this system our clients get all the benefits of a modern electronic kanban system without the significant investments in infrastructure.

Try it for Yourself

Schedule your 1st Free Consultation and Start your AI Journey Today