In digital manufacturing, legacy systems still loom large
If you think legacy systems are your biggest hurdles toward Industry 4.0, we’ve got some advice for you.
Four years ago, a pilot smart factory sponsored by the German Research Center for Artificial Intelligence collaborated with chemicals manufacturer BASF SE to produce fully customized shampoos and liquid soaps.
A test order was placed online that generated radio identification tags, which were attached to empty soap bottles on an assembly line. These tags communicated to production machines on the floor what kind of soap, fragrance, bottle cap color, and labeling each individual bottle required. A wireless network linked all of the machines and products to each other to facilitate a fully automated workflow where only the machines and products exchanged information and filled the order. The only thing a human did was to key in the original order.
The pilot established a paradigm for Industry 4.0, which aims to automate manufacturing production processes with Internet of Things (IoT) technology. A central key to making pilots like the soap project work is being able to combine the efforts of a legacy order entry system with the workflow of IoT devices and label. The goal is to take over the process from the point that orders are entered, and transformed into finished, customized products without any further manual intervention.
Unfortunately, four years later, many companies still struggle with their efforts to convert to smart manufacturing that uses both IoT and legacy systems.
A recent survey of 39 manufacturers we recently conducted confirms this.
In the survey, we found that:
82% of survey respondents said they wanted to integrate IoT data with legacy systems.
35% of these survey respondents said they only had one to three systems to integrate to achieve legacy system-IoT communications, but 23% said they had four to six different systems to blend, and 30% said that they had over ten different systems to integrate.
“The reality is that almost everyone has some sort of legacy system to deal with, whether it’s machines of various vintages and conditions, an MES (executive manufacturing system), or an aging AS/400 midrange server,” said Karen Field, Executive Director of Content for the Penton Internet of Things Initiative.
It’s equally noteworthy that legacy systems aren’t going away anytime soon. Some ofthe world’s largest ERP system installations run as legacy systems and for good reasons: they are perceived to be mission critical to enterprises, and they are painful to rip out and replace.
The challenge, then, for manufacturing executives and CIOs, is how to effectively transform manufacturing operations with IoT automation that not only coexists with, but that leverages the value of legacy systems?
Here are some recommendations:
Define precise manufacturing goals
What are your manufacturing goals? Is it to customize a bottle of soap on the fly so time to market can be improved and maximum customer satisfaction can be gained? Or do you want to automate machine-issued alerts to let maintenance staff know when to repair machines before they fail so production downtime can be eliminated? IoT and manufacturing automation goals should be clearly defined and agreed to upfront by stakeholders so everyone sees the value.
Identify your system integration points
How many systems do you need to integrate that results in an end-to-end IoT workflow? You might need an order entry system to auto-trigger an order fill that initiates IoT action in your manufacturing plant, or you might need a combination of order entry, inventory, purchasing, robotics, and sensor-based IoT systems to interact with each other for ERP purposes. All of these systems and data integration points should be clearly identified and assessed for integration complexity before you begin to build a project task list and timeline.
Keep project sizes small
If your goal is to automate an entire manufacturing plant, don’t try to do it all at once. Instead, lay your projects out piece by piece. The first project might be to insert robotics on an assembly line to automate part of the assembly build process. A second project might be adding a machine to perform a visual inspection of goods as they move down the production line.
What you want to achieve is a step-by-step success record of IoT implementations that deliver tangible value to the business (i.e., reduced time to market, cut costs, more defect-free products, less time lost to production floor downtime due to equipment failures, etc.). This chain of successes allows you to build your case for smart manufacturing powered by IoT.
Prepare your data
If your purchasing department knows a particular supplier as “AXE Company,” but your manufacturing department only knows the supplier as “Axelrod,” how do you ensure that both departments know they are communicating about the same supplier? You can’t say this unless you standardize the supplier under a single uniform data name that everyone uses, and you link all of the related alternate names for the supplier to one standard name.
It isn’t uncommon for many companies to achieve uniform data naming process by hand. In fact, 32% of HULFT survey respondents said they were consuming significant staff time cleaning and preparing data, and 27% said that they had to resort to manual entry of data as part of the data cleaning, preparation and entry process.
Can we just be honest?
At HULFT, we believe in having open conversations with our customers about the realities of how legacy systems fit into their digital manufacturing journey. For all the hoopla about Industry 4.0 and the IoT era, most of the customers we talk to are still trying to figure out how to close the gap between where they are now and where they want to be 5-10 years from now. Is that you? If it is, we want to hear from you. Our team will work with you to problem-solve within that gap, set realistic metrics, and help you and your stakeholders assess how quickly you’re reaching your goals.