“We’d love to do more with AI, but we can’t find the people.” I hear this from manufacturing leaders constantly. The skills gap is real—but it’s not an excuse for inaction.

There are practical strategies for bridging the gap. Some involve finding people. Others involve developing people. Some involve restructuring how you approach AI projects entirely.

The nature of the gap

First, let’s be specific about what skills are actually missing.

Data engineering: Getting data from factory systems into formats that AI can use. Connecting PLCs, historians, SCADA systems to modern analytics platforms.

Machine learning/AI development: Building and training models that work in industrial contexts. Understanding what’s possible, what’s practical.

Domain expertise + digital fluency: People who understand manufacturing processes AND can work with data and AI tools. This combination is especially rare.

Change management: Getting frontline workers and supervisors to actually use AI systems. Technical skills are necessary but not sufficient.

Vendor evaluation: Knowing enough to evaluate AI vendor claims critically. Too many companies buy snake oil because they can’t assess what’s being sold.

Different projects need different mixes of these skills. Not every AI initiative requires a PhD in machine learning.

Strategy 1: Hire differently

The traditional approach—post a job ad, wait for the perfect candidate—doesn’t work well in a talent-short market. Try:

Hire adjacent skills and train

Someone with a controls engineering background can learn data analysis faster than a data scientist can learn factory operations. Look for people with automation experience, electrical engineering, or process engineering backgrounds who are curious about AI.

One Newcastle manufacturer hired a process engineer who’d taught herself Python on YouTube. Six months later, she was building anomaly detection models that outperformed vendor offerings.

Target career changers

Mid-career professionals switching from IT, academia, or other industries bring valuable skills. They might lack manufacturing specific knowledge but can learn it—especially if they’re motivated.

Consider remote/hybrid for specialist roles

Data scientists and AI engineers can work from anywhere. If you’re in regional Australia or can’t compete on salary with Sydney/Melbourne tech companies, remote arrangements expand your candidate pool significantly.

Work with universities

Co-op placements, internships, and capstone projects can build a pipeline. Students get real experience; you get to evaluate potential hires over months rather than hours of interviews.

RMIT, UNSW, UQ, and several other universities have specific programs connecting AI/data science students with industry. The good students often have multiple offers, so build relationships early.

Strategy 2: Upskill existing staff

Your existing workforce knows your operations intimately. That knowledge is incredibly valuable—and much harder to teach than AI skills.

Identify high-potential employees

Look for people who:

• Show curiosity about technology
• Have figured out workarounds using spreadsheets or basic tools
• Ask “why” about data and processes
• Have credibility with their peers (important for change management)

Provide structured learning paths

Random YouTube tutorials aren’t enough. Structure matters. Options include:

• Certificate programs from universities (many now have industry-focused AI/data courses)
• Vendor certifications (Siemens, Rockwell, etc. have training programs)
• TAFE courses in automation and data analytics
• Online platforms like Coursera, edX (but with supervision and application to real problems)

Apply learning immediately

Training without application is wasted. Pair learning programs with actual projects. “Here’s a Python course” is less effective than “Learn Python and build a dashboard for this problem we actually have.”

Create time for development

This is where many companies fail. People are too busy with their day jobs to learn new skills. If AI capability matters, carve out dedicated time—a few hours a week, or dedicated project time.

Strategy 3: Partner strategically

You don’t need to build all capabilities internally. External partners can fill gaps—but partner relationships need to be structured thoughtfully.

Use consultants to build capability, not just deliver projects

The worst consultant engagements leave you with a working system but no ability to maintain or extend it. The best build internal capability alongside delivering results.

When engaging consultants, insist on knowledge transfer. Pair your staff with their experts. Document everything. Plan for the transition from consultant-led to internally-led.

Consider hybrid teams

Some companies use a permanent core team supplemented by contractors or consultants for specialist skills. This provides flexibility and access to expertise without the cost and risk of hiring specialists you may not need long-term.

Evaluate vendor support levels

If you’re buying AI software, understand what support comes with it. Some vendors provide extensive implementation assistance, training, and ongoing support. Others hand you software and wish you luck.

For teams with limited AI expertise, strong vendor support is worth paying for.

Strategy 4: Restructure your approach to AI

Sometimes the answer isn’t finding more skilled people—it’s making AI projects require less specialised skill.

Choose platforms over custom development

Building custom AI from scratch requires serious expertise. Using pre-built platforms for industrial AI (Siemens MindSphere, AWS IoT, PTC ThingWorx, etc.) requires less.

You trade some flexibility for accessibility. For many applications, that’s the right trade.

Start with augmented intelligence, not automation

Tools that help humans make better decisions are easier to implement than tools that make decisions autonomously. You don’t need perfect models if a human is validating outputs.

Pick problems carefully

Some AI problems are genuinely hard. Others are straightforward once you have the data. Focus early projects on achievable wins. Build capability and confidence before tackling ambitious challenges.

Buy before build

Custom development sounds appealing but requires ongoing maintenance, updates, and expertise. Off-the-shelf solutions for common problems (predictive maintenance, quality inspection, energy management) are often good enough and much less demanding on scarce skills.

The realistic outlook

The skills gap isn’t going away soon. Australian manufacturing is competing with tech companies, consulting firms, and overseas opportunities for the same limited talent pool.

But manufacturers who take capability-building seriously, who create environments where skilled people want to work, and who partner effectively can still make progress.

The companies I see succeeding don’t wait for perfect hires. They develop people, partner strategically, make smart technology choices, and build gradually. It’s slower than having an unlimited talent pool—but it works.

Waiting for the skills gap to resolve itself is a losing strategy. The gap might narrow eventually, but your competitors aren’t waiting either.