2023 ASEE Annual Conference & Exposition

Data analytics and Artificial Intelligence (AI) have transformed many industries in the last decade. In tandem, a skilled workforce needs to understand how to gather/access data to extract trends and optimize operations, and how to label the key events and develop training data sets which can be used by machine learning (ML) experts for advanced analytics. The power of ML and AI has not been fully realized in the manufacturing sector. One of the major challenges is that the small and medium manufacturers which account for 98% of industry lack the dedicated data analytic workforce. This is combined with aging workers and significant challenges in hiring factory floor workers.
To address this need, partnerships have been established between industry and academia through Wabash Heartland Innovation Network (WHIN) at Purdue University. In collaboration with Ivy Tech Community College, a series of workshops were developed to introduce data analytics, internet of things and basic machine learning concepts to local small and large manufacturing companies. This study will describe three short courses geared toward industry workers and professionals. The first short course is on the topic of energy savings and data analytics for Variable Frequency Drives (VFDs). The main goal of this workshop was to introduce electric motor data and VFDs for motor control to industry partners. Motors are a major source of energy consumption in manufacturing and other industries. In right applications, VFDs can reduce energy usage with relatively short return on investments. Using VFD or utilizing SMART motor overload devices, it is possible to gather data for motor diagnostics, predictive maintenance, and process monitoring. The second course is on the topic of recording and managing data from factory workers’ observations. A key requirement for some machine learning application is the training data set and labeling of key events to complement automatically recorded machine data. In this workshop, AirTable will be introduced as a method to reduce paperwork and help capture key events and observations. Examples from electric motor maintenance and computer numerical control (CNC) machining are provided. The third short course will be on the topic of industrial internet of thing (IIoT) sensors. The course objective is to introduce IIoT device installation, accessing the data (e.g. power consumption, vibration, temperature) and use dashboards and alarms to monitor the operations.
To provide formative feedback on the impact of the three short courses for industry on data in manufacturing, we used an adapted framework of the Kirkpatrick Training Evaluation Model (Kirkpatrick, 1996), combining elements of Guskey's Evaluation Model of faculty development (Guskey & Sparks, 1991). Kirkpatrick's and Guskey's models have the first two levels (Level 1 and Level 2) in common, namely, assess of participant's reactions and learning. The third level (Level 3) focuses on organizational support for promoting change, accounting for contextual effects. The fourth and fifth levels of the model (Level 4 and Level 5) focus on the use of the new knowledge and skills as applied in practice and the overall impact on the organization.
This study reports on participants' perceptions regarding Levels 1, 2, and 3 of our evaluation model. Levels 1 and 3 were assessed immediately after the workshop through a survey, and level 2 required applying a form of assessment immediately after the workshop (i.e., performance task). In our future work, we will assess Levels 4 and 5 annually, starting a year after the implementation of the short courses, using objective metrics and interviews with managers. This assessment strategy will start over for each cohort of new participants entering the professional development program.