Verista is highlighting its COUNTQ Tray Inspection System, an AI-enabled inspection and reconciliation system designed to count and verify tray-loaded products in pharmaceutical production environments.
The system uses deep learning AI, high-resolution imaging, LED lighting, and electronic verification to improve accuracy and throughput compared with manual counting. It is intended for products including vials, syringes, ampoules, and other tray-loaded parts where reconciliation errors can create rework, delays, and compliance risk.
Verista says the system can deliver electronic verification of product counts in 15 seconds. COUNTQ is aimed at applications where rule-based vision tools are inadequate, using AI-enabled inspection to manage variation that would be difficult to handle through conventional image thresholds alone.
The platform also supports compliance requirements, including controls relevant to 21 CFR Part 11 environments. Pharmaceutical inspection systems must do more than produce a correct count. They must generate records that can be reviewed, audited, validated, and defended within regulated quality systems.
Tray reconciliation remains a stubborn production problem because it often sits between automated manufacturing and human quality assurance. Manual count checks can be labour intensive, ergonomically poor, and vulnerable to fatigue. Automated systems can reduce that burden only when they integrate cleanly into validated workflows and handle product variation without creating new exception loads.
Inspection becomes a data problem
Pharmaceutical manufacturers are under increasing pressure to improve throughput while maintaining strict quality control. Production lines are expected to handle higher demand, more complex product mixes, and tighter documentation requirements. Inspection and reconciliation systems therefore have to support both operational speed and regulatory confidence.
AI-enabled inspection changes the way manufacturers approach that balance. Traditional rule-based vision tools work well where variation is limited and image conditions are predictable. They are less effective when products, trays, reflections, fill states, labels, or component orientation create visual complexity. Deep learning systems can classify patterns more flexibly, provided they are trained, validated, and governed appropriately.
Validation is the hard part. In pharmaceutical manufacturing, AI cannot be treated as an unexplained black box sitting between operators and release decisions. Manufacturers need evidence that the system performs consistently, that changes are controlled, that records are secure, and that exceptions are managed in a way quality teams can approve.
That quality and manufacturing context is visible across Europe’s pharmaceutical production base. Sandoz’s Slovenian biosimilar capability shows how advanced medicines manufacturing depends on technical development capacity, process discipline, and scalable quality systems. Inspection and reconciliation tools such as COUNTQ operate in the same environment of rising complexity.
The logistics layer is also becoming more technical. Temperature-sensitive medicines, speciality pharmacy shipments, and validated distribution workflows require structured decision-making, as shown by digital tools for pharmacy shipment risk and pack-out selection. Manufacturing and distribution are becoming more data-rich from line clearance to delivery.
COUNTQ’s operational appeal lies in reduced manual effort and faster reconciliation, while its quality value comes from electronic evidence, repeatability, and reduced discrepancy handling. The business case will depend on labour availability, product value, line speed, rework rates, and the cost of delayed release.
The technology also reflects the direction of pharma automation. Companies are automating the smaller checks that sit around high-speed filling and packaging steps, because those checks often determine whether a batch moves smoothly through the plant or becomes trapped in investigation and documentation.
COUNTQ’s success will depend on integration as much as accuracy. The system has to fit into existing production layouts, validation regimes, operator routines, batch records, and quality procedures. If it does, AI-enabled tray inspection can reduce friction in one of the more repetitive but consequential parts of pharmaceutical manufacturing.
