The saying "time is money" may be considered metaphorical, but for companies involved in shipping goods (e.g. fulfillment centres for the e-commerce industry) it is very accurate. There, every second is valuable as it means more orders can be fulfilled and every shipment sent means additional revenue. However, companies face a problem because, while buying goods online can be done virtually with a few clicks, delivering them requires a complex warehouse and logistics infrastructure.
The first step in this very non-virtual process is picking, which is finding the goods in the warehouse and getting them to a location where they can be packed for shipment. This may seem like a fairly simple process, but imagine a warehouse the size of several football pitches, thousands of orders to be processed per hour and hundreds of staff, scanners and various types of trolleys dedicated to picking. It is this scale that makes combining the individual orders that we want to deliver quickly, without errors and efficiently a major challenge. It is a very difficult organisational as well as... mathematical problem. Why? Because it can only be well addressed with data, as shown by the experience of apps like Uber, which also need to efficiently "deliver" many different passengers to different places.
How does this logic work in the warehouse?
With up-to-date information about the distribution of goods, orders and knowing the current location of employees or forklifts involved in the order picking process, it is possible to optimise the process of allocating orders to pickers, just as the Uber algorithm does with drivers. Every improvement in the average order completion time of an order sent by a customer, translates into real benefits, as it allows you to fulfil more orders per day with the same resources. The larger the scale, the greater the benefits, especially as the e-commerce industry grows and shoppers expect deliveries to be faster and faster. This is particularly important during peak shopping periods, such as before Christmas. This is when fulfillment centres are under more time pressure and can't afford delays in deliveries. Automatic picking systems using pallet stacker cranes and roller conveyors are increasingly used in distribution centres. Not only do these systems allow for an increase in the number of picks, but they also offer major savings in operating costs by making better use of space and reducing rising labour costs. In traditional warehouses, picking often accounts for more than 50% of costs.
Despite high implementation costs, automation certainly pays off in the long term, as it speeds up average picking times and reduces labour costs. Unfortunately, it is not always possible, as it depends on the type and dimensions of the goods handled, the layout of the warehouse and the space available. In the majority of distribution centres, the automated section therefore only occupies a fragment of the warehouse space and it is the workers, usually using various types of trolleys, who pick the orders. It is for this part of the warehouse that our solution is applicable, because in the case of warehouse space numbering in the tens of thousands of square metres and the number of resources involved in picking counting in the hundreds of employees, effective management of the picking process and resource planning is impossible without the help of IT systems. Our customers are also looking for ways to improve order picking processes because it is becoming increasingly difficult to find qualified employees, so optimisation is also considered in the context of not only reducing labour costs, but also caring for the comfort of the already employed staff.
"Uber for warehouses".
Indoorway's CEO, Adam Komarnicki, draws on his experience from implementing the solution in over 30 production and warehouse halls of companies such as: LOTTE Wedel, Whirlpool, Solaris, Dr Irena Eris, Adient. The Indoorway system is based on the UWB (Ultra Wideband) technology, which works similarly to GPS, but allows for very precise location measurements of resources moving inside buildings. In the case of distribution centres, such assets are mainly trolleys, employees and scanners. By knowing their exact location in real time, you can send orders to the best-located asset and manage the entire picking process to optimise picking times and distances, while keeping in mind the expected shipping times for each order. Real-time asset location information also allows you to react quickly in the event of picking delays, e.g. if a scanner involved in a particular picking has been put down or a cart has been idle for too long.
Similar to Uber applications, by seeing on a warehouse map where all the trucks, workers and scanners are at any given time and how they are moving, we can optimise each picking order so that the worker or truck does not have to travel unnecessarily long distances, especially 'empty', i.e. before it reaches the shelf with the goods. This optimisation increases overall warehouse efficiency, allowing more orders to be processed for customers. Looking at the cost side, process optimisation lowers the average cost of picking, reduces the number of resources involved and allows for better space utilisation. For e-commerce order processing operators, continuous process monitoring allows for better control of the service level agreement (SLA) with the corporate customer.
Saving time and money
Komarnicki emphasises that such a solution is needed because, at present, warehouse managers have no information on where individual stocks are at any given time. The only available information is the place of the last warehouse operation registered with the scanner. This is potentially useful information, but can be misleading - after all, an employee may have put the scanner down in the meantime or moved to a completely different location, which will not be registered in the WMS. The Indoorway solution, with its continuous location measurement, records exactly what happens between scans.
Location data is a very good complement to WMS records because, in addition to being able to better manage the picking process in real time, it also adds context to individual transactions, which helps identify sources of problems and better plan resources. The combination of both types of data will show where bottlenecks are created that delay individual picks, how much the picking paths differ from those determined by the algorithm, what the use of warehouse space looks like for different types of goods, how many resources are actually involved in picking and how much time these resources spend productively.
Summary
The use of location data is already common in logistics, but is only now entering the more micro level, that of warehouses. Such information helps to address problems that cannot be solved in a traditional way, e.g. a very well-known problem in mathematics called the traveling salesman problem, and in logistics called the marshalling problem. The aim is to determine optimal transport routes for a given number of means of transport, which are to serve designated customers located in different places. These types of optimisation algorithms have already been tested by companies such as Uber, among others, but only recently can they be applied to accurate location data for warehouse traffic. And here we come back to the point - optimised single order times translate into more orders served, which in turn translates directly into higher revenues, and at no additional cost. In the logistics industry, where there is very high pressure on prices, operational efficiency is a powerful competitive advantage, and at this point there is no better way to achieve it than to rely on modern technologies and advanced data analytics.
Want to find out more? Visit the Indoorway solution page and see where the technology for asset movement location and process monitoring is performing well, or email us at indoorway@aiut.com