AI to transform warehouses
10 June 2021
Five years ago, I was asked a seemingly simple question by the owner of SEC Storage... are we relevant? One month of furious research later, I gave my answer: ‘Yes, but we won't be for long’ before outlining an ambitious plan to revolutionise our approach.
This included revamping our warehouse design ability by developing an Artificially Intelligent Design-Bot called ELSA to work alongside our expert design team.
Five years, five national awards, and unprecedented growth later, ELSA has helped transform our company into an industry-leading provider of warehouse solutions and has demonstrated the power AI has to supercharge distribution facilities regardless of size or industry.
So, what does ELSA do? ELSA's role is to leverage our clients' data to ensure Everything Lives Somewhere Appropriate. ELSA utilises order, product and volumetric data, alongside cutting-edge machine-learning techniques, to find optimal clusters of 'similar' Stock-Keeping-Units (SKUs). This process is essentially a sophisticated form of ABC analysis and classifies products that should be stored in similar types and sizes of Pick-Faces.
Our designers then work with ELSA to refine these classifications by analysing the behaviour of each SKU and simulating virtual environments based upon seasonal data and growth projections to suggest the correct types and sizes of pick-face that balances pick-efficiency and replenishment cost.
The benefits of ELSA: In a typical warehouse, order picking and replenishment typically equates to more than 50% of the total cost of the facility, of which half is travel time. Pick-face selection and sizing is, consequently, the most critical design decision in most warehouse operations. Unfortunately, traditional pick-face selection regularly results in over-sizing pick-faces due to an understandable but over-inflated fear of escalating replenishment costs. This has two negative impacts: first, pick-face density is reduced and travel distances increase. Secondly, oversized pick-faces lead to excessive "white space" and larger aisles, resulting in reduced capacity.
Mathematically speaking, pick-face selection is an example of an 'optimisation problem', the type of query which machine learning is specifically equipped to tackle. ELSA's removes the overcompensation bias of traditional design methods and human designers and ensures that every SKU is analysed and classified scientifically.
How effective is ELSA? Since her introduction, ELSA has demonstrably improved all of the core performance indicators we use to assess the quality of our designs. Our solutions are now more effective and efficient, reducing return on investment periods substantially, which in turn has supported a 376% increase in the rate at which our customers order our proposed solutions. For illustration, SEC's award-winning project for Focus International increased pick-efficiency by over 100% and provided a 40% increase in capacity within the same footprint.
Major external bodies have also independently validated ELSA's ability to improve design. Since ELSA's inception, SEC has won four and been finalists for over ten other national awards for warehouse design, innovation and efficiency. Furthermore, last year, SEC was awarded the Lloyds Bank National Business Award for Data Excellence due to our disruptive approach to warehouse design.
ELSA has, unquestionably, transformed our business by providing SEC with the ability to revolutionise our clients' operations. Our data-driven approach works regardless of industry sector or business size and is supplied to genuine potential customers free of charge. So if you are considering how to transform your warehouse operation, reach out for a free, no-obligation consultation today.
Harry Watts, managing director, SEC Storage