From since the 1950s automated guided vehicles have been used in the warehousing environments. With the recent improvements in Artifical Intelligence (A.I) such as localization and mapping, this has allowed the guided vehicles to move across traditional manufacturing boundaries and travel between buildings. These improvements have also led to the creation of warehousing software such as warehousing execution systems (WES).
Lowe’s is one example of AI being used in warehouses. Their autonomous retail robots would help customers and create real-time data by scanning their inventory by looking for patterns in product and price discrepancies.
Predictive analytics is another use for A.I in warehousing. This feature is created by IBM, which collaborated with the Weather Company so that they are able to predict the weather from newsfeeds and social media to predict likely conditions. This prediction can allow for retail owners and managers to make better strategic decisions on supply, pricing and promotions. For example, when the weather is dry and hot, hamburger sales tend to increase, therefore hamburger meat will be promoted during this period.
Thirdly, reducing waste is one of the key benefits of implementing machine learning into the warehouse management. Picking in-stock items is vital for reducing waste. One way to do this is that the workers will pick the items first including the bin rotation, serial and lot number tracking. After this is done, machine learning adds in the intelligent planning. This means it can suggest alternative lot to ship, based on history/need of the product.
Lastly, when it comes to the delivery aspect of the product machine learning excels at visual pattern recognition, which makes it vital for the task of physical inspection and maintenance. This system is able to identify if a shipping container or product were damaged, time of damage and solutions to repair the damages. This includes visuals, reports and recommendations in real time.
In the Supply Chain, Artificial Intelligence can (and is) wildly used by a lot of giants such as Amazon for their warehouse technology. Not only they allow to be faster, but they also make the company save money on the long-term.
In a more general way, when it comes to warehousing, AI can be used for:
- Improving inventory accuracy
- Supporting decision-making
- Being a cheaper and quicker workforce
Machine learning can have a huge impact on data collection and decision making, as a warehouse produces on a daily basis huge amounts of data collected from order numbers, inventory stock levels and shipping data. Machine learning can learn about the data through patterns and algorithms and therefore suggest activities such as the restock of a product when that one is almost out of stock.
This can learn a warehouse employee’s voice at specific places where they are in the warehouse and then confirm the accuracy of inventory records.
Robots can replace humans for some warehouse tasks, where they exceed them in speed and treated orders per hour. The company Ocado, in one of their old warehouses, is able to process around 3,5 million items or 65,000 orders a week. They do basic tasks such as “lifting”, “moving” or “sorting”. The bots are geared with claws that they use to grab crates which they pull in their central cavity. Then they move the crate to another location and drop it.
It would be wrong to say that the robots are ‘intelligent’, or at least to mean it individually. In fact, their actions are all coordinated by a central computer. They can however help each other to go pick a quite unusual item that may be some place far, and accomplish in a few minutes a process that can take hours in a more traditional warehouse. Since all the robots are based on the same model, they are all interchangeable. Which means that if one breaks, it can be easily replaced. This results in economies of scale, as the mechanical diversity is drastically reduced.