You are here

Future Vision: Increasing Production with Predictive Technology

How software tools, models will forecast future behaviors, outcomes in laundry

CHICAGO — Most laundry and linen services are constantly searching for ways to increase production, to get more pounds per operator hour, while maintaining quality, and ensuring goods that are needed are done on time.

Jon Witschy, sales manager for Spindle, a technology company that builds infrastructure for industrial operations to measure and improve productivity, talked about increasing laundry production with predictive technology at the Association for Linen Management’s (ALM) IMPACT conference in November during the session, “Alexa: How Can I Increase Production?”

Predictive technology is the use of tools, such as software tools and models, to monitor and analyze patterns to forecast likely future behaviors or outcomes, says Witschy.

The data comes from many sources, such as sensors or counters on laundry equipment.

“In simplest terms, we’re taking what we know, we’re running it through these analytic tools and we’re developing insights for what we what we need to do next,” he shares. “The concept is straightforward, but the tools that are needed are relatively complex because they’re kind of replacing the brainpower and the judgment and the decision-making that we have but moving it into that model.

“However, think of how much faster we could make such determinations. Think about how your PC or a calculator can perform addition, subtraction, multiplication; how much faster your computer can search through volumes of data to get information. Imagine applying that speed to the decisions that you have to make every day in order to run your business.”

Laundry technology today has often been (and can still be) “descriptive,” collecting information that describes what has happened.

“Thus far, even for systems that are collecting and presenting data in real-time, that information is still only descriptive in terms of the story of what’s happened up to this point,” says Witschy. “For example, I can determine that I processed this number of sheets thus far today either from a manual log or a counter or control system on my finishing equipment.”

The next step would be being prescriptive, which he describes as “kind of the first step in moving toward predictive.”

Prescriptive technology mines the data and uses analytic tools to be able to identify some causes and outcomes, perhaps even combining multiple sets of information to generate recommendations of how to address the current situation.

“As an example, I could find that I’ve produced enough pillowcases today to meet my daily delivery requirement, but that I’m short on sheets,” Witschy says.

“So the prescriptive tool would tell me make an adjustment, switch my ironer from small pieces to large pieces, move my team around—using the information I’ve collected up to now to tell me what I should do right now.”

Being predictive, he says is “foretelling the future” in a way. The technology puts the collected data into a model, and then the model helps an operator plan ahead.

“Continuing on my sheet example, I know that I’ll need X number of sheets tomorrow, and the system will analyze and know the capacity and performance of the machines, what will meet that requirement, and it could even inform me that operator A is my best feeder,” he says.

“So, that’s who I want to put on that machine to get sheets out tomorrow and thus get the best performance out of my operation.”

The move to predictive technology involves the latest evolution in industry, known as either the Internet of Things (IoT) or Industry 4.0, Witschy shares.

Industry 1.0 involved initial mechanization to help speed up industry. Industry 2.0 moved into mass production and assembly lines and utilizing electrically driven components. Industry 3.0 is where industry started to implement robotics, electronics, sensors, etc., to become better at collecting information.

“But now, finally, we’re having the ability to network disparate systems, the ability to access information remotely and even affect something that’s going on back in the facility when you are away, and then the software tools to analyze all of this information and help us make informed decisions or looking forward perhaps even makes those decisions for us,” points out Witschy.

“Imagine equipment on the floor providing more analytic tools and decision-making capabilities, to have the equipment be able to take that next step so that there’s no operator or manager intervention to what’s going on.”

He points out that the return on investment (ROI) for predictive technology comes from bringing multiple systems together and being able to, at a glance, see exactly what’s going on in all areas of a facility, what’s being delivered to a plant, etc.

“For me it’s just the transparency of information, making it accessible, certainly putting it online,” he says. “I can access any one of my customers’ facilities and see what’s going on in their plant right now from my cell phone, as you can do with just about everything, putting it on multiple interfaces.

“Or how about just being informed of the conditions that would indicate or preclude a catastrophic failure or even just, hey, we need to step up the preventive maintenance schedule on this because we’re hitting the edge of the envelope there a little too soon.

“Certainly we still need to have experts in certain areas, but we don’t have to necessarily pull in that expert for every minor decision that takes place in the plant because we can have these software tools that are going to help us do that and point us in the right direction.”

Witschy says that predictive analytics is “where things start to get really exciting.”

“We’re really going to hit the transformational point in our industry,” he says. “This will come down to taking that information again, correcting our current actions, as well as taking steps to make sure that we don’t repeat them in the future or be ready to ask when those next instances of may occur.”

Witschy notes that the technology is still at that point where humans need to give more input back into the systems to help them continue along a “maturity model path.”

“I hope that everybody is interested, intrigued and excited about the future with prescriptive and predictive technology, as it is getting a foothold in our industry,” he concludes, “to continue to develop out there and look forward.”


Artificial Intelligence: Future of Laundry Operations? (Part 1), Dec. 17, 2020

Artificial Intelligence: Future of Laundry Operations? (Part 2), Dec. 22, 2020

Artificial Intelligence: Future of Laundry Operations? (Conclusion), Dec. 24, 2020