This article is submitted by the MECHANICAL ENGINEERING, MANUFACTURING and ENTREPRENEURSHIP Unit, UTT

Manufacturing in Trinidad and Tobago has always been defined by resilience. To remain competitive in a global market that continues to evolve, the sector must keep building value, improving efficiency, and adopting technologies that support growth. One such technology is Artificial Intelligence (AI), which has moved from being a novelty to becoming a strategic necessity. For manufacturers in Trinidad and Tobago, the question is no longer whether AI matters, but how local firms can adopt it in practical, affordable ways that deliver real results on the plant floor.

This was the context in which UTT's Mechanical Engineering, Manufacturing and Entrepreneurship (MME) Unit partnered with KC Confectionery Ltd. The objective was to move beyond theory by first delivering foundation AI training, and then supporting the company in developing pilot solutions aligned to its specific manufacturing challenges.

AI is often presented as a "black box": expensive, complex, and only accessible to large multinational companies. However, the most effective starting point for many local manufacturers is internal capability building.

UTT MME delivered structured training at KC Confectionery focused on practical plant realities. The sessions covered data basics, how AI models learn, and the importance of clean, reliable data. Importantly, the training also focused on decision-making: selecting the right problems, defining success metrics, and planning implementation in a way that supports production rather than disrupting it. Following the training, KC's team identified three practical solution areas to pursue as pilots.

The first area was visual anomaly detection. Quality losses are often caused by small deviations that accumulate into waste or rework. While manual inspection remains important, it can be inconsistent at high production speeds. KC explored an AI-assisted vision approach using camera input to flag unusual patterns. At the pilot stage, the focus was on identifying relevant defects and improving model performance in KC's actual plant environment.

The second area focused on moving maintenance from reactive to proactive. Unplanned downtime is one of the most costly forms of waste in manufacturing. Although many firms maintain maintenance logs, this data is often underused. This pilot explored how historical maintenance information could help identify recurring issues and support preventative maintenance planning.

The third area was data-driven cost efficiency. Cost reduction requires clear visibility into where losses occur. Factors such as energy usage, material waste, and changeover time can be difficult to isolate without structured analysis. KC's cost-efficiency pilot used operational data to identify high-impact improvement opportunities and support evidence-based continuous improvement.

Pilots are an important first step. They prove concepts, build confidence, and help teams understand what is required for wider implementation. However, moving from pilot to full deployment requires stronger infrastructure, including reliable data pipelines, validation across shifts and conditions, clear governance, and integration with existing systems.

AI implementation comes with challenges, including data quality, workforce readiness, and change management. However, these should not prevent manufacturers from starting - they do not need perfect conditions to begin. They can start with focused pilots, learn quickly, and build the local capability needed to scale AI responsibly.

For further information on UTT's Mechanical Engineering, Manufacturing and Entrepreneurship programmes, visit www.u.tt/mme, call 642-8888 ext. 32157 or email nianna.alexis@utt.edu.tt

Photo caption: Dr Nadine Sangster, Associate Professor and Programme Leader, Mechanical Engineering, Manufacturing and Entrepreneurship, UTT (back, right) and Dr Jorrel Bisnath, Assistant Professor, Mechanical Engineering, Manufacturing and Entrepreneurship, UTT (back, second from right) with employees of KC Confectionery Ltd during a session on applying AI solutions to their manufacturing process.