Artificial Intelligence (AI) is quickly moving from an emerging technology to a practical tool for card manufacturers. That shift was reflected in recent conversations at ICMA EXPO, where industry leaders discussed the growing role of AI, as well as in feedback from InBrief readers who expressed interest in learning more about how AI can be applied within the card industry. One area where AI may offer significant potential is the manufacturing process itself. The following framework explores how AI can be incorporated into quality management for dual-interface card production to help manufacturers identify risks earlier, prevent defects and improve production performance.  

Dual-interface card manufacturing leaves little room for error. In a production environment where small variations in milling depth, bond pressure, antenna orientation or certification status can compromise performance, traditional quality controls may no longer be enough. As card programs become more complex and production volumes increase, manufacturers need ways to identify risk earlier, prevent defects before they occur and build continuous improvement into the production process. 

That is the focus of a framework developed by ICMA member Siva Ravi, Director of Quality Management at Marketing Card Technology LLC. Ravi outlines how card manufacturers can combine ABCXYZ risk analysis, AI-enhanced Failure Mode and Effects Analysis (FMEA) and AI-driven Plan-Do-Check-Act (PDCA) to move from reactive defect detection to proactive prevention in T-Connect dual-interface card production.  

Why Early Detection Matters 

In dual-interface card production, the cost of poor quality grows quickly as each card moves through the line. A defect caught during inlay gathering may cost as little as 15 cents per card, but that same defect can cost more than $10 per card, plus network fines, if it is not caught until after field issuance, a more than 67-fold increase in cost. 

According to Ravi, this escalation is especially important in what he describes as the “critical risk zone,” which includes inlay gathering, hot stamping, T-Connect milling and bonding. In these stages, an invisible inlay orientation issue, failed hologram adhesion, a milling deviation of just 0.03 mm or a small bond pressure variation can create major downstream losses. 

The lesson is clear: every undetected defect that advances through production carries more embedded value and greater financial exposure. By applying AI-driven, 100% inline verification earlier in the process, manufacturers can interrupt that escalation before defects become expensive scrap or noncompliant cards. 

Mapping External Risk Before It Reaches Production 

Not every production failure begins on the factory floor. Some of the most disruptive issues originate upstream, including supplier variability, late artwork approvals, inconsistent data files, changing chip specifications or expired certifications. 

Ravi recommends using ABCXYZ analysis to classify those risks before they affect the production line. The ABC portion ranks jobs by revenue contribution, helping manufacturers identify where the financial stakes are highest. The XYZ portion ranks demand predictability, including how consistently external inputs arrive. 

The most dangerous combination is AZ: high-revenue work with unpredictable inputs. These programs require a more disciplined approach, including predictive material staging, tighter verification and stronger compliance gates. Ravi notes that AI can help manufacturers stage high-value materials only when a production run reaches a high probability of moving forward, reducing waste and limiting exposure. 

One overlooked risk is chip certification expiration. If a program moves forward with an expired certification, the issue can become a major compliance problem after cards are already produced or issued. Ravi argues that manufacturers are often in the best position to prevent this failure because they have both the chip knowledge and the operational visibility needed to catch the issue early. 

AI-Enhanced FMEA for Structural Integrity 

FMEA is already a familiar quality tool for manufacturers, but Ravi’s framework applies AI to make it more powerful in dual-interface card production. Traditional FMEA evaluates process failures using Risk Priority Number (RPN) based on severity, occurrence and detection. The higher the RPN, the greater the priority for corrective action. 

The challenge is that traditional sampling methods can miss clustered defects. Inlay orientation errors, milling depth deviations and bond failures may occur in patterns tied to equipment condition or material lots. If only a sample is inspected, some of those defects may escape. 

AI-enhanced FMEA changes the detection model by applying 100% inline verification. Instead of relying on statistical sampling, manufacturers can inspect every card for critical attributes such as inlay orientation, milling accuracy, T-Connect bond quality, Q-factor performance and hologram placement. Ravi’s framework shows that improving detection in this way can sharply reduce RPN across major failure modes. 

For dual-interface cards, this matters because the product amplifies small errors. A slight milling deviation can detune the antenna. A contact interface issue can cause terminal rejection. A contactless performance drift can create failures that are difficult to detect until the card is used in the field. 

AI-PDCA: Building a Learning Quality System 

Ravi also extends the traditional PDCA cycle into what he calls AI-PDCA. While the classic Plan-Do-Check-Act model supports continuous improvement, AI can help the cycle become more predictive. 

In this model, AI systems sense production signals, reason across data, act through guided interventions and learn from results. Instead of waiting for defects, audits or complaints to reveal a problem, manufacturers can use AI-enabled systems to identify patterns earlier and adjust before failures spread. 

This approach supports a more resilient production environment, especially for complex card programs with tight tolerances, variable inputs and high compliance requirements. 

A Case for Competitive Advantage 

Ravi cites a real-world global bank dual-interface card program involving 2.5-5 million cards, custom holograms, precision milling requirements and tightly controlled inlay specifications. The production environment included unpredictable approvals, inconsistent data deliveries and last-minute engineering changes. 

By combining predictive pull through ABCXYZ analysis with AI-enhanced FMEA and 100% inline verification, the program achieved significant improvements: a 40% reduction in work-in-progress inventory, an 85% drop in scrap, an increase in SLA compliance from 65% to 99.2%, an 80% reduction in approval time and a 90% decrease in reliance on supervisor intervention. 

For card manufacturers, the larger message is not only about quality. It is about competitiveness. In a market where price pressure is constant, the ability to prevent defects, manage volatility and protect compliance can become a defining advantage. 

The New Standard for Smart Card Manufacturing 

Dual-interface card production has raised the stakes for quality management. Traditional tools remain important, but they must evolve to match the complexity of today’s programs. ABCXYZ analysis helps manufacturers understand external risk. AI-enhanced FMEA strengthens structural quality. AI-PDCA turns continuous improvement into a data-driven learning system. 

Together, these tools can help manufacturers move from finding defects to preventing them. For an industry built on trust, security and precision, that shift may be essential to the future of card production. 

Why ICMA Membership Matters   

ICMA remains committed to empowering the global card manufacturing community by helping members lead, innovate and grow. Through a strategic combination of industry marketing support and professional development, ICMA equips organizations and individuals with the tools they need to stay competitive in a rapidly evolving marketplace.  

ICMA’s professional development programs are designed to strengthen both technical expertise and leadership capability. Offerings such as on-demand Advanced Card Education (ACE) and the Card Industry Training & Education (CITE) initiative help professionals deepen their knowledge, while flagship events like the EXPO and CardTREX bring together industry leaders to learn, connect and explore emerging innovations.  

Members who engage in ICMA’s education programs and events don’t just keep pace with change—they help shape it. From contributing thought leadership to Card Manufacturing magazine to earning industry recognition through the Élan Awards of Excellence, ICMA provides meaningful opportunities for members to elevate their visibility and influence.  

For card industry professionals seeking to stay informed, connected and future-ready, ICMA offers a clear path forward through education, exposure and community.  

Discover the full value of ICMA membership and join a global network of professionals advancing the card industry.