Machine Learning in SME Warehouse and Logistics Management

Machine learning has really taken great strides in recent years. Your phone can tell you to leave for work early because traffic is bad and homes are starting to fill with voice activated smart assistants. In the area of warehousing and logistics machine learning is helping companies to reduce costs, increase efficiency and improve accuracy.

Machine learning can help a logistics business to more effectively distribute consignments across drivers to improve delivery speed and increase the number of deliveries per driver. Route optimisation algorithms help a driver to avoid traffic and get to the destination sooner.

Using machine learning to predict supply and demand, a warehouse can better manage the distribution of items across locations to reduce movements, improve picking and replenishment speed, reduce congestion and maintain more accurate stock counts.

In the office machine learning can be used to improve the speed and accuracy of data entry, or in many cases it could eliminate data entry entirely. The user interface could automatically adjust to match the company’s workflow without having to be explicitly configured to do so.

Large companies have the edge in machine learning right now. They have the funds to invest in the technology, and they have the data to feed the hungry learning machine. They also have a position of power, being able to force others to comply with their rules, shifting the cost and burden onto someone else’s balance sheet.

Small to medium sized companies typically cannot afford to invest in technology at the same level and operate with much lower numbers of items and operations, meaning the amount of data for machine learning is considerably less. There is also a different requirement in that they often need to comply with rules and regulations dictated by larger companies.

The challenge for CartonCloud is to adapt the machine learning methods of larger companies so that they better suit the requirements of small and medium operators, while dealing with a more limited and diverse data set. As a first step in this direction, we’re increasing our data gathering capabilities so that we can begin delivering optimisations via machine learning to give our customers a competitive advantage.