By Petra Mitchell, President and CEO, Catalyst Connection
Southwestern Pennsylvania has the manufacturing experience and technical expertise to compete for more business, but recent growth has lagged comparable regions. Technology can help manufacturers close that gap by enabling their existing workforce to increase output and take on more business. To achieve those gains, manufacturers can begin with the operating need, put artificial intelligence (AI) to a practical test, make employees part of the investment, connect better performance with new business, and measure the next step.
The Pittsburgh Regional Alliance (formerly the Allegheny Conference on Community Development) State of Manufacturing in Southwestern Pennsylvania report illustrates the gap. Regional manufacturing generated $15.3 billion in gross regional product and supported more than 90,000 direct jobs in 2024. Yet manufacturing output grew 6.3% between 2019 and 2024, compared with approximately 21% in both Cleveland and Charlotte.
Industry mix appears to explain a substantial part of that difference. Regional chemical manufacturing output grew 77.6% over that period, compared with 6.1% for primary metal manufacturing, which accounts for a larger share of manufacturing output here than in Cleveland or Charlotte. These differences point to the need to build on existing capabilities while pursuing higher-growth markets.
For the region’s technology community, this creates an opportunity to work with manufacturers on improvements tied to factory-floor conditions and customer needs.
For an individual manufacturer, that effort starts with identifying what limits output, margins, or new business. Before considering automation or AI, company leaders should ask whether equipment capacity, workforce shortages, or customer requirements for quality and traceability are restricting growth.
Workforce availability is a frequent obstacle, according to Matt Minner, Managing Director, Technical Services, at Catalyst Connection. Understanding its effect on operations helps companies assess shop-floor automation, equipment condition monitoring, and systems that connect production information.
“Technology is not always the right lever,” said Minner. “An assessment should examine operations, costs, and available data before a company selects a tool. It can guide a phased investment plan and clarify what a technology provider must deliver. Both parties should agree on the scope of a pilot and how its results will be evaluated.”
Production data might reveal why a machine is idle or where scrap is reducing available capacity. Digital quality records could help employees identify recurring defects. Equipment information may support more timely maintenance. Each application needs a baseline so the manufacturer can determine whether the change improves performance enough to justify the investment.
For example, an iconic Pittsburgh bakery deployed SCADA and MES systems to replace paper-based shift reporting, reducing production-data delays of 6 to 12 hours or more to real-time visibility. RFID tagging now tracks product location throughout production, strengthening traceability and helping the bakery meet customer and audit requirements needed to access new markets.
Once the operating need is clear, manufacturers can choose the simplest useful tool. AI’s value depends on the task, available data, and expertise needed to make its output reliable.
“AI is just one portion of the technology toolbox,” Minner said.
Frances Phan, Data & AI Specialist at Catalyst Connection, sees interest in AI for administrative tasks, quoting, work instructions, and analyzing large amounts of information. Measured business results from some projects are still pending. Manufacturers should distinguish an application worth testing from one that has demonstrated a return.
A simpler approach may be sufficient. Phan described integrating separate Excel files into master files without a more complicated AI application. A dashboard connecting existing information may likewise provide the visibility a manufacturer needs.
“You need good data,” Phan said. “Paper records and disconnected systems can leave employees working with incomplete or delayed information. Before introducing AI, a manufacturer needs to understand where its information comes from, whether it is reliable, and who is responsible for maintaining it. Even a company with integrated core systems may have separate activities that require attention.”
One emerging application involves extracting dimensions and tolerances from engineering drawings to prepare inspection forms. AI may reduce manual effort, but quality management professionals must verify the output. A pilot should measure accuracy and time savings, including the effort required to review and correct the work.
A useful first project has a defined scope, an accountable owner, and employees who understand the tool’s capabilities and limitations. Before expanding its use, the company should compare results with the original baseline and determine whether the improvement holds up under everyday working conditions.
AI can also deliver measurable time savings. A Lawrence County industrial component manufacturer uses AI-assisted coding to develop custom modules for its MRP platform, giving the company capabilities it likens to having “30 developers on staff” and enabling rapid iteration. The customizations save two employees a combined 20 hours of manual work each week.
Selecting a useful tool also requires input from those who will work with it. Regional manufacturing employment declined 3.3% between 2019 and 2024 while productivity rose, underscoring the need to help employees work effectively and develop skills that support growth.
“People, buy-in, and engagement are the difference,” Minner said.
Operators, supervisors, and maintenance personnel should help define requirements and test proposed systems. They understand production conditions and can identify obstacles that equipment selection alone may overlook. Their participation helps a successful pilot become part of daily production.
That involvement is especially valuable when introducing digital work instructions to support newer employees. AI may help prepare documentation or make information easier to find, but capturing experienced workers’ knowledge still requires interviews and human interaction. On-the-job training and time to adapt procedures remain essential.
Training should address actual tasks and decisions. Measures such as output per labor hour, scrap, unplanned downtime, or time needed to learn a task can show whether employees are benefiting.
Responsible AI use also calls for a policy defining sensitive information, permitted tools, and required human review. Phan emphasizes attention to customer and industry requirements. A designated group can assess proposed applications and determine whether they are producing useful results within those requirements.
Operational improvements support growth when they help manufacturers win and fulfill new orders. Companies should identify prospective customers’ specifications, certifications, traceability, and delivery requirements before investing in processes or technologies.
Experience in metals, machinery, electronics, or chemicals may be relevant to energy and other markets. Certifications, consistent quality, and dependable delivery can affect whether a company qualifies. Minner also identifies cybersecurity requirements as a consideration for manufacturers pursuing defense sector work. Those requirements should be understood before resources are committed.
Dependable suppliers are essential to fulfilling those commitments. Manufacturers can map critical materials and lead times, use purchasing records to identify recurring shortages, qualify additional sources where specifications permit, and share forecasts with important suppliers.
The region’s 2,800-plus manufacturing locations in 2024, including more than 630 fabricated metal products locations, offer a substantial base for supplier relationships. Finding the right partner requires assessing quality, capacity, and delivery performance as well as proximity.
With customer requirements understood, manufacturers can direct resources toward demonstrating their capabilities and fulfilling new orders. Digital quality records can support customer evaluations, lower scrap can improve order economics, and productivity gains can create capacity. Reliable suppliers and well-trained employees help companies meet their delivery commitments.
Over the next six to twelve months, manufacturers can identify one growth constraint or promising market, establish a baseline, and test a focused improvement. Measuring results can guide further investment and show whether to expand the application, adjust it, or address a different need.
The State of Manufacturing report offers useful perspectives on the region’s strengths and growth gap. Closing that gap depends on individual companies matching their capabilities to customers’ needs. Manufacturers and technology providers can work together on practical improvements that help companies compete for and deliver new business.
Petra Mitchell is President and CEO of Catalyst Connection, a Pittsburgh-based not-for-profit organization that provides consulting and training services to small manufacturers in southwestern Pennsylvania.