Ready to jump-start a digital manufacturing strategy? Ten steps to ‘crawl, walk, run.’
By Jennifer Clement and Noel Hopkins
On the 28th of November 2023, 100+ senior executives gathered in Green Bay to discuss “Investments in AI and Digital Technologies to Address Labor Challenges.”
A panel of four senior executives of manufacturing and logistics companies headquartered Northeast Wisconsin shared insights regarding how their AI and digital strategy has evolved.
The panel was made up of diverse industry sectors: T1 automotive supply, transportation/logistics, packaging, and food manufacturing. As each shared how they are experimenting with technologies, guests got a peek into how AI is transforming the way things are made and moved around the globe.
Early in the discussion, however, it became clear that many in the audience were not as far along as the panelists. “How do we go from Excel to advanced analytics?” one participant asked.
In this article, you’ll get 10 steps to help crawl/walk/run as you consider your own digital journey.
As you think about some of your key business challenges and find out how others are solving them, feel free to reach out to CLA to discuss a Digital Readiness Assessment.
Step 1: Start simple – take a class.
“Encourage a team member to take a class on data visualization to learn how to interact with dashboards. That’s not AI – it’s industry 1.0.”
Step 2: Automate a great process vs. a bad one.
“We used to solve problems by adding people and those days are over. At one point our use of temps swelled to 150 people in one facility. As we explored solutions, it became clear that we first had to address dated and deficient workflows. We used process mapping to make waste visible, see where we were entering data twice, and streamline steps. Then we wrapped automation around a better process. We reduced 150 temps to 20 and we’re still chipping away at it. Failure is the beginning of success.”
Step 3: Divide investments in three pillars.
“Our approach involves three pillars. The first is about keeping the lights on. The second is about core strategic apps that differentiate the business. The third involves emerging technology transformation. A year ago, we were investing just 1% on transformation. Today we’re up to 20-30% and we’ve made a lot of progress.”
Step 4: Build excitement around expanding career paths.
“Initially there was fear that automation would eliminate jobs. We assured we are not replacing people with AI. AI enables us to explore whole new career paths. AI and humans will compliment each other. AI allows people to work on more value-add activity.”
“We asked each operator, ‘which job do you like the least?’ and we automated them one by one. So far, we’re seeing hundreds of hours of savings. Robotic Process Automation (RPA) was our first step to automate manual steps and we needed operator input from the get go.”
Step 5: Choose a ‘Proof of Concept’ with real ROI
Example #1
“Key to our process is controlling moisture – too much or too little has a major impact on costs. In the past we relied on operators making gut decisions to regulate moisture. Today we’re experimenting with operators interacting with sensors and machine learning. So far we cut costs related to controlling this process in half. We also added gamification. Every day operators are ranked from 1-5 on their performance. We take their feedback to tweak the model and get even better accuracy.”
Example #2
“Customers are pickier than ever and demanding higher quality. Designing dies is complicated. There are three major challenges: we don’t have enough people, our process is really complex, and troubleshooting is even more complex. Instead of guesswork, we’re using machine learning to find trends that help reduce downtime.”
Example #3
“AI is taking over help desk and customer service – we will continue to see more.”
Example #4
“We used to use Optical Character Recognition (OCR) to convert manuals and SOP binders to text. AI goes beyond OCR which just reads characters. AI understands context.”
Example #5
“In the logistics business, we need data in real time in order to make our process more efficient. We collect data from thousands of drivers and also trucks to enable predictive analytics. Waiting 1 month or even 1 week for data is too long. We created a ‘data lake’ to enable staff to use ‘generative AI’ to ask natural questions about real problems. It used to take us weeks to build custom reports. Now it takes seconds.”
Example #6
“We started with skilled technicians to be more data driven. With 40 years of experience comes 40 years of guesswork at the machine level from your top operators. And now with sensor technology dropping in price, it’s easy to get data from machines. First we framed up goals around profitability and then pulled the data we need drive decision making. We had to make correlations to prove the data works and take emotions out. It’s a game of science.”
Step 6: Standardize data.
“Our acquisitions keep coming and today we are juggling seven different ERP systems across multiple service lines. Our difficulty lies in standardizing data, which makes the difference between being in the red or black. We had to make sure everyone calls things the same thing. We hired a third-party data scientist which is a fast and affordable way to clean up your data and enable your resources to scale.”
Step 7: Learn from failures.
“Early on we spent on a vision system with no clear ROI. The idea was to use machine learning to target intermittent recurring defects. But – the defects were so intermittent that there wasn’t enough data for the machine to learn. Our lesson learned: be intentional and confirm the applications can really solve the problem.”
Step 9: Deploy security.
“Data is an asset and we manage it like one. We don’t lock it down but protect it within corporate boundaries. Early on we created policies for who has access, how they access, and what they access.”
“We utilize a tracing system to look through network traffic and see who is working off hours, uploading large quantities to external web servers, etc.”
“About a year ago we realized our internal security team was not doing enough. So we partnered with 3rd parties to help keep watch. Within a couple of weeks they identified 30X more threats more than we did on our own. This approach allows our small three-person team to do more strategic work. With security a moving target, it makes sense to outsource to others who can keep up with it better than we can.”
Step 10: The bottom line – crawl, walk, run.
All panelists agreed an intentional shift mindset was step zero followed by small steps in pockets to build momentum. “I’ve been with company 20 years and our IT team pivoted from reactive to proactive 12 months ago. Today we focus on the art of the possible. And we’re only in the first inning.”
“We had to change our mindset from perfection to progress.”
Need help jump starting your digital roadmap?
Start with a Digital Readiness Assessment, and CLA’s professionals can help with that. Let’s talk.
Jennifer Clement is an executive sales and marketing leader specializing in value creation for the C-suite. In her current role at CLA, Jennifer collaborates on strategy with executives of global manufacturing and distribution companies to accelerate results. Previously Jennifer served as a Global Business Acceleration Leader for Complete Manufacturing and Distribution (CMD). During her time with CMD, Jennifer lived and worked in Asia from 2015-2019. Prior to CMD, she spent 10 years in senior care technology. Jennifer started her career at Johnson Controls (JCI) and spent nine years in leadership roles; followed by five years at Rockwell Automation (ROK) leading c-suite strategy and marketing operations.
Comments are closed.