What is the role of technology in warehousing? Most of us think we already know the answer, and it has to do with robots sliding around warehouse space in the dark, picking orders and generally acting like little autonomous vehicles. However there’s another way that warehouse operators are using technology – and that’s to respond to customer demand.
Automated decision making has two different forms:
- software recommending to an individual that they take an action or decision
- software automatically taking an action or decision based on available information.
Let’s look at that in detail as it’s an important part of the process to understand.
In the first instance, software might collate the number of orders arriving in a warehouse and suggest to 
the warehouse manager that they organise another order collection from their logistics provider. In the second, the software will be calibrated so that at a certain number of orders, it automatically contacts the logistics provider to request another or an augmented collection.
Why consider automated decision making?
When supply chains are under constant pressure, warehouse managers have to innovate as part of the demands to have competitive advantage. Both logistics and warehouse management processes can be streamlined and optimised through automated decision making.
Most automated decision making includes two elements – artificial intelligence (AI) and machine learning. These can seem mysterious, but essentially the way they work together is that AI is the facility a computer has to see patterns in data and machine learning then ‘decides’ whether the pattern that has been identified is positive or negative – whether it leads to a success or win or a failure or loss.
This is significant because more data arrives in a system than human cognition can factor into decision making. So, for example a human will consider current weather conditions, planned public holidays and staff rotas, while automated decision-making might also factor global weather patterns, international holidays and health patterns around the world, to explore how supply and demand might impact a warehouse, both within its own space and through a range of external factors that simply don’t register with humans. Automated decision making also tends not to be human-centric – which can lead to more imaginative thinking and less reliance on human resources and their inherent limitations.
What is warehouse automation, in general?
Any combination of mechanical processes (conveyors, stackers, cranes, lifters used for pallet storage etc) and software (including data collection and decision making) can be considered automation. Automated processes should deliver one of these three attributes
- Better range
- Great agility
- Faster speed.
This means that automated warehouses processes should, by definition, produce more activity with fewer resources. This is particularly significant with omnichannel order fulfilment where order origination and last mile delivery may both by highly varied.
What warehouse decisions can be automated?
While some decision processes can be automated, others don’t respond so well to an automated process. For example – ideal automated decision-making processes include:
- when to release orders
- who work should be assigned to in a warehouse
- the optimum route for order picking.
In essence, every warehouse activity is a process and those processes break down again into two levels – the first level is the decision making process where information has to be processed to create the best activities, using the fewest resources. This involves day to day decisions – which orders to prioritise, which logistics provider to use, how to pack an item and much longer term decisions such as how to set up a warehouse system and where to store each SKU to deliver the best efficiency level in picking orders. The second level is the actual product handling process where decisions are made about how to move items around the warehouse, from arrival, to storage, to pick and pack, to outbound logistics.
On both levels, automated decision-making offers advantages – on the decision level there is less risk of human error and a greater level of efficiency based on data rather than opinion or previous experience and on the product handling level, automated decision-making also reduces exposure to human error but also increased warehouse efficiency because quicker processes also increase the potential for faster and more satisfactory order fulfilment that eventually delivers greater customer satisfaction.
Where is your warehouse today?
There are four levels of potential automation in any warehouse setting:
a – low or no automation – especially common in small warehouses this is a system that has almost entirely manual processes or very basic automation with no automated decision making.
b – systems automation – the most common position in small to mid-sized warehouses which is were some form of automated decision making exists, usually through the implementation of a Warehouse Management System (WMS). It may also be possible to link this WMS to basic automation equipment such as RFID and voice picking.
c – automated goods handling – relatively common in the UK, this is the level of automation where automated decision making is integrating to the point where it permeates the warehouse at several levels such as convey belt systems and automated retrieval of SKUs.
d – when the full functionality of WMS is implemented in a warehouse space, complete automation of the decision-making process can influence warehouse activity at all levels including but not limited to automatic sorters, automated guided vehicles (AGVs), robotic picking systems and automated palletisers for bulk orders. This level of automated decision making, allied to automated product handling, is largely limited to big, often purpose-built, warehouse facilities such as Amazon.
How to assess your need for automated decision making
Obviously, whether you’re looking to begin automating a warehouse or considering increasing your current level of automation to implement automated decision making, the basic considerations are always the same.
- What are the specific requirements, limitations and possibilities of your warehouse operation – which can be established through a SWOT analysis.
- What are the potential gains to be achieved through automated decision making – which can be established through market research and talking to suppliers of automated decision-making systems.
- What are the costs inherent in upgrading your warehouse space and what ROI can you expect.
There is always the option to explore warehouse rental by relocating to warehousing that has inbuilt automated decision-making systems.