Not all explosions are bad!
The ‘information explosion’ may seem like a new term, but it was first used in the 1960s, for instance in advertisements on IBM’s data storage. And the information explosion continues today. Researchers at IDC have predicted that by 2025 we will be generating 163 zettabytes of data per year. To explain: 1 zettabyte is 1 trillion gigabytes!
Does that sound scary? I don’t think so. The more data becomes available, the more opportunities there are to improve processes, optimize planning and predict the future. This ocean of data adds value to your business, by enabling you to make better choices and decisions.
It’s all about anticipation
It’s very handy to be able to predict the future, particularly if you need to plan routes or staffing. An accurate prediction can save you lots of money. Take transport, for instance. The further in advance you can tie down your capacity, the cheaper it will be. The same applies to staffing. If you know long beforehand how many people you will need, you can start your recruitment campaigns in time to attract the right people, rather than having to hire expensive temporary staff at the last minute.
The more you incorporate data in your analyses, the better your predictions will be. It’s impossible for a person to keep an overview of all the tiny details. But that’s where computers, calculation models, algorithms and machine learning come in. They make data clearer and easier to understand.
You can take better decisions by letting today’s data guide your actions tomorrow. Not based on intuition or experience, but based on facts. Take this practical example.
A logistics firm in the ornamental horticulture sector is asked to make lots of extra runs during the course of the day, even though its schedules have already been drawn up. This means making lots of last-minute arrangements. And trucks that were almost back at the depot have to turn and go back to pick-up points that they’ve already visited. Highly experienced planners might eventually develop a sense of when to let trucks wait for potential shipments, but that’s the exception, not the rule.
By analyzing historical data it has proven to be possible to make predictions about when and where return shipments might arrive. This can be taken into account when planning the drivers’ routes. Drivers can be proactively instructed to wait in a certain region where a truck is going to be needed at any second. That reduces unnecessary mileage, increases capacity utilization rates and improves customer satisfaction.
Here’s another practical example: in retail distribution it’s often important to arrive on time. Many stores and distribution centers operate with time windows that have to be met.
By analyzing historical data, it is possible to identify certain patterns that a planner would have more difficulty seeing. For instance, there may be a certain store where unloading takes a significant amount of extra time on certain days. This can be taken into account when planning the schedule for the rest of the day, so that time windows can also be met later on. It is also possible to find out whether the cause might be at the shop’s end. In any case, this means improved service for customers later in the day and possibly process improvements at the store in question.
Taking data into account
Both examples show how historical data analysis can be applied in practice. So just imagine how proactive your planning could be if you take into account external data as well, such as weather forecasts and news (e.g. about strikes in a customer or supplier’s organization).
Achieve better results in 3 steps
In any case, data-driven logistics (Dutch article) starts with identifying the persistent problems in your logistical processes. Then you establish what data is available. The final step is to combine logistical expertise with data science.
Would you like to find out more? Click here to contact us.