How to Optimize Material Flow with Data Analytics and Machine Learning?
As a manufacturer, you probably know what to expect from the future. You will have to produce more customized products and run smaller lot sizes while increasing production velocity and reducing costs. Ultimately, this will test how efficiently you can manage the material flow. Are you ready? Many manufacturers are not because of one simple reason: siloed data.
This article explains how you can break down the silos and optimize material flow easily with data analytics and machine learning.
Download the Material Flow Optimization whitepaper with two customer cases!
You can ask any manufacturer, and they see a complicated and unpredictable industrial landscape ahead. Yet, one thing is sure: consumerism continues its evolution.
In fact, consumerism is one of the key trends behind the blurry future outlook in manufacturing. Seasons tend to become shorter, more regional flavors are demanded, and buyers favor unique products. Money is spent on cheap, short-life consumables increasingly.
Despite the rising consumerism, people do not necessarily settle for poor quality. Also, products must be manufactured in a sustainable and environmentally friendly way, and higher transparency and social responsibility are required.
How does consumerism impact manufacturers?
Well, they must produce more short-run batches and change over lines from producing product A to product B more frequently. In addition, the ever intensifying competition implies that manufacturers must continuously reduce production costs and shorten lead times. The traditional make-to-stock model does not fly anymore – manufacturers must align with just-in-time operations. These are just some of the reasons why material flow optimization becomes critical in the future.
Given the above, are manufacturers ready for the challenge? Yes, if they can transform their operations using lean practices and technologies – an efficient material flow lays the foundation for sustained competitive advantage!
What’s material flow in manufacturing?
In manufacturing, material flow means transportation of raw materials, components, work-in-progress inventory, and final products through the production chain – from inbound to outbound.
In an optimized flow, materials at different stages move steadily, and as predicted, through the system. It can scale up and down based on fluctuating volumes, and there are no unplanned stops and starts in the process.
Benefits of optimized material flow
A smooth material flow reduces your transportation, storing, and handling costs, as well as capital investments. Also, it minimizes waste, energy consumption, and the need for floor space. As a result, resources and work time can be used efficiently. And ultimately, you can shorten your production time.
Essentially, an efficient material flow shows as an increased bottom line.
In the future, however, optimized material flow accounts for more than just profits: you can comply with increasingly stringent sustainability and transparency requirements, reduce carbon footprint, and use scarce raw materials more efficiently; in other words, you can run production more sustainably.
Why is material flow optimization so difficult?
Because of siloed data!
In any manufacturing organization, there are several stakeholders involved in the material flow: procurement, material planning, production planner and manager, sales, finance, and fulfillment. Often, you can find these teams working in their silos; they have siloed processes, databases, systems, and dashboards. Needles to say, the systems do not communicate across the department borders.
If you take a closer look at their production processes, you can see that the machines produce a lot of valuable data, but it is difficult to collect due to various machine-specific formats and interfaces. Alternatively, it might be that the a manufacturer is able to collect data, but due to lack of platform that can truly leverage these diverse data sources, the data “goes to waste”.
The silos prevent end-to-end visibility to the material process. So, how can you transform this for the better?
RELATED READING: Break Down Data Silos To Increase Efficiency
How to optimize material flow: the 4 stages
Building the platform for an optimized, end-to-end material flow, which facilitates efficient working, is a vast undertaking. That’s why only a handful of manufacturers have been able to transform.
The following points give a quick run-through of the four stages of material flow optimization to make the transformation easier. You can download our whitepaper for a detailed description.
1. Collect and Centralize Data
Do you know why in many manufacturing organizations, the operators and managers, still today, must access several systems to get the needed information? The root cause is siloed data and systems that do not talk to each other. This results in a massive waste of time and potentially increases the likelihood of human errors.
The first step is to break down the silos and enable end-to-end access to data. For this, data must be collected along with the whole process – from all possible data sources – to a central location such as a virtual data lake.
This might sound trivial, but remember that data is collected from various different sources – sensors, warehouses, inventories, enterprise systems, execution systems, and PLCs.
CUSTOMER CASE: Elisa Smart Factory helped a global consumer goods manufacturer to harmonize its data from more than 20 systems. Download the whitepaper for more details.
2. Automate process monitoring
Centralized data is only the beginning. Now, you must put this data into productive use.
You can do this by establishing continuous, automated monitoring and rule-based notifications for the material flow to ensure that it goes as planned. The programmed rules detect process deviations – such as bottle-necks, material shortages, or overflows in the in-process inventory. When a rule recognizes that a threshold is exceeded, it automatically raises an alert and sends out a notification to operators who can take the appropriate actions to e.g., prevent bottle-necks, storage overflows, or material shortages.
3. Visualize real-time data
Now the automated monitoring runs in the background and makes sure your process is under control. As a result, the operational staff and managers can lean back and concentrate on their work.
With visualization, you can give the operators and managers tools to carry out their tasks in the most efficient way. This includes intuitive situational views into the end-to-end material flow, showing only the information that is relevant in each role.
You have two options: a full-blown 3D digital twin, and a simpler, performance dashboard.
The 3D digital twin shows what is happening in the factory at any given moment so that operators and management decisions can be based on real-time data. A digital twin can cover an entire factory, or one, or several production lines. With a simple click on an asset, you can quickly drill down to see the full process and machine-level KPIs.
The Performance dashboards provide easy to read, one-page summaries of the real-time material flow status. They combine all the relevant information for each role and give an overview of the material flow at a glance. And, when needed, you can quickly drill down deeper into the details.
CUSTOMER CASE: Learn how a fast growing electric car manufacturer benefits from consolidated data intuitive, role-based performance dashboards. Download the whitepaper for full details!
4. Predict and Act in Advance
You have now built up a full situational awareness of the material flow. This is fine today. In the future, you must know in advance what’s going to happen in the material flow.
Why? Because production velocity increases, batch sizes decreases, and change-overs must be carried out even more frequently and faster. Accurate just-in-time deliveries are a must.
So, you must be a few steps ahead of the process to have enough time to react to sudden incidents and fix problems on time.
This requires Predictive Analytics and Machine Learning. It means that you can teach computers to detect anomalies and data outliers, which have previously anticipated an incident such as a bottle-neck or storage overflow. When you apply these algorithms in the real-time data flow, you can predict incidents and raise alerts earlier. You can now have more time to act before the process is affected – e.g., you can replenish the inbound storage in time.
You can download our Material Flow Optimization whitepaper for a detailed solution description and two example cases.
The benefits of material flow optimization boil down to one crucial capability: end-to-end visibility to the entire material flow instead of seeing data in isolated silos. It’s the key to massive cost savings, higher operational efficiency, better quality, and improved productivity.
Elisa Smart Factory helps global manufacturers such as Procter & Gamble, Danfoss, and NVision to solve demanding industrial challenges. Our team of industrial data scientists is ready to work with you for a smoother end-to-end material flow – Contact us and let’s get the transformation going!