Higher Education faces challenges in providing learning experiences and cultivating competencies in our modern environment. Academic discourse centres on tutoring automation, a complicated educational dilemma in the age of Machine Learning, Deep Learning, and Artificial Intelligence. Establishing a structured technique for real competency assessments is an early challenge. Therefore, this research develops and implements a structured evaluation process to assess student competencies in complex problem-solving scenarios. The study finds dimensional variation in student performance through statistical analysis of multivariate data from two engineering courses, demonstrating that each evaluation method measures different competencies. Key findings demonstrate that while traditional evaluation methods provide quantitative results, the newly proposed instruments offer a nuanced understanding of student abilities by capturing quantitative and qualitative data. This approach standardises the assessment process and lays the groundwork for developing intelligent tutoring systems to automate and personalise learning. This method may lead to computer-assisted instruction to adapt educational technology for engineering complex problem-solving and other skills. It provides a solid foundation for personalised learning technology assessment methods.