When Jasmeet Singh, a thought leader in the data and analytics space, took on a new role for the global hydraulics manufacturing company HUSCO in Pewaukee, Wisconsin, the company presented him with the daunting task of transforming the I.T. organization from a reactive support group to become a key business driver with focus on digital manufacturing. Jasmeet found a significant gap in organizations’ desire to use cutting-edge technology and the skills I.T. possessed to support it.
“When I finished the assessment of the current state of systems, processes, and talent, I realized how big of a challenge I was dealing with. Almost all of the I.T. systems were close to being the end of cycle. Additionally, no trust in the I.T. organization existed as users exhibited frustration with slowness and regular downtime. Plus, the I.T. department possessed outdated skills.”
Where To Start?
How Jasmeet tackled this was by focusing initially on people. He brought in accomplished talent from outside and picked the right partners to fast-track key initiatives. In addition, Jasmeet built a strong training program to upskill existing talent. He highly recommends putting your leadership team in place and having strategic alignment with the business. Do this all before you start shopping for new software. The new talent they brought to HUSCO created an immediate impact by improving existing systems and processes. This was low-hanging fruit that helped build trust with the business right away.
“One of the major reasons for production downtime was that data drives will run out of space. HUSCO had an archiving process in place, but no one was watching alerts of archive process failures. They had too many alerts enabled. So, just by doing a simple exercise of cleaning up the alerts and putting service level agreements in place to take action resolved the issue.”
“Once basic things were taken care of, we had to prepare to support the company’s exponential growth. As a supplier to the world’s major automotive companies, HUSCO is required to have 0 defective parts per million with no delays in delivery,” Singh stated. “The company needed a holistic look at the entire end-to-end process. Not just what was inside the boundaries of our manufacturing plants. We needed to integrate our entire supply chain and adopt all aspects of Digital Manufacturing.”
What is Digital Manufacturing?
Digital Manufacturing, according to Jasmeet, is when all your systems and processes are integrated end to end. This includes everything from taking a new sales order to issuing a production order. It includes creating a purchase order to receiving raw materials, assembling the parts, testing each part for quality, packaging, shipment and delivery, parts calibration at customer sites, and much more.
HUSCO had transformed its assembly and test stands to use robotics and automation. This meant they were collecting millions of data points every second. These needed processed, analyzed, and sent in real time to other systems. So, the moment a part leaves the test stand and the serial number is assigned to it, they needed to update the production order number and sales order number to kick start downstream processes. To meet on-time delivery and quality, HUSCO needed to introduce some rigorous processes at its suppliers. Yet, the suppliers did not have all the advanced capabilities they had in-house.
Based on a scale of data and the need to operate globally, Jasmeet knew that HUSCO needed to adopt a cloud-first strategy. They started by implementing a cloud data lake using Azure and later upgraded our custom applications to Azure Web Apps. This way, applications became externally accessible with a high degree of security. Using Azure Active Directory, these applications were available externally and enabled our suppliers to send subcomponent data in advance as all they needed was the Internet. They had hundreds of suppliers, so providing the same access to the same application also made sure they received data in a consistent format. This helped eliminate post-processing and any manual changes.
“Next, we started integrating all other business systems like Product life cycle management, quality management, enterprise resource planning, etc. with our manufacturing execution system and also with using APIs,” Singh said. “For the systems which did not support APIs we chose a backdoor way of updating the backend databases directly. After the integrations completed, we started upgrading and migrating each of these business systems to the latest and greatest including our E.R.P.”
After laying the foundation of data lake and clean consumable data, they leveraged Machine Learning, Artificial Intelligence, Augmented Reality, Robotics Process Automation, etc. to make all areas of business predictive and more efficient.
This was no easy feat. There was a lot of resistance on the way because HUSCO was not just upgrading systems and processes, they were introducing tremendous change which was too much for a few folks, also some of the initiatives like cloud-first. At first, R.P.A raised a lot of eyebrows and skepticism even within I.T. but once they overcame that resistance by educating and proving the value what HUSCO ended up with was truly amazing.
End Results for HUSCO
For the first time, HUSCO had one manufacturing dashboard. This enabled them to view productivity, utilization, first time through, etc. globally for each production line in real time. Plus, it gave them the ability to drill down to the root cause. They had their M.E.S system send an alert as soon as a part failed quality check. Or, even before test metrics started deviating in the wrong direction. They had the capability to start and stop production lines remotely or even automatically based on data.
“Our supply chain analyst no longer needed to prepare a scorecard manually each week for hundreds of suppliers as it was delivered automatically,” Singh said. “Supplier data exchange application that onboarded all suppliers meant that we no longer had to wait for parts to arrive at our dock before testing them for quality, then returning them if our system will reject them. This would happen automatically as soon as the supplier uploads the required data.”
HUSCO had M.L models predicting raw material quantities more accurately and also optimizing shipping frequencies so neither end up with too much raw material blocking shelf space nor did they run out of parts needed to meet our production targets, ordering the right amount of material at the right time also meant savings in holding cost and also getting best rates for shipping which helped us survived inflation and tariffs.