AI as Collaboration Platform in Engineering Classroom
Published by : Industrial Automation
Dr Harshal Oza elaborates upon how Artificial Intelligence can provide a collaboration platform for bringing Industrial Automation to Engineering Classroom.
Manufacturing in India, and particularly in the SME segment, is under-automated for various reasons. One urgent imperative to improve this situation is to expose engineering students to industrial automation. It is always a challenge for Engineering Faculty to strike a balance between engineering science and engineering practice. Interestingly, both industry and academia have their own and different methods for training the student. There are many approaches by Industry and Academia to collaborate, with varying degrees of success. One way of achieving this is by using AI or Artificial Intelligence as an aid to assist the process of teaching and learning. By using AI as the medium of collaboration, we propose an effective model where the two realms can overlap for the benefit of students, faculty and their prospective employers. A solution, named ‘ed up me’, built by AptRaise Technologies is introduced.
The aim of this article is to discuss a model that bridges the gap between the industry and academic activities. The article also introduces a solution that implements the discussed model.
1. Automation as an implicit skill
a. Inherent challenge: multidisciplinary requirements
Automation has become an essential part of manufacturing around the world due to ever increasing demand for fast turnaround times, tighter tolerances, and expectation of higher productivity. Incorporating automation in a process or a plant requires participation from multiple disciplines. However, the adaptation of automation technology remains a multidisciplinary challenge for several MSME units in India.
b. Challenge for industry: Shortage of skilled engineers
One of the biggest challenges for SMEs of India is that of recruiting and skilling. The skill requirements for automation are spread across many engineering disciplines. Automation technology essentially requires knowledge of sensors, actuators, robots and computing in the context of automating a plant or process. It is a must to see automation as a design problem with awareness on data, communication and security.
c. Challenge for universities
It is very important that the academic institutions impart on students the concepts of automation technology so that they become fluent with the language of automation. However, this needs to be achieved with the current availability of skilled faculty and resources.
2. How is Industrial Automation taught currently in academia, training institutes, industry and their assessment methods
a. Course on Automation – Universities
The main aim of academic institutes is to give a strong foundation to students of engineering. The flavour with which the course is taught is dependent on the discipline. For example, it is common to find a course on Robotics offered by Mechanical Engineering departments to have a heavy bent on kinematics, dynamics and design of robots. A department of Computer Science may offer it with emphasis on artificial intelligence and machine learning. In some institutes, Industrial Automation includes the basics of PLCs, Drives, Hydraulics and Pneumatics packed into one six-month long course.
b. Third party training institutes
The main aim of these institutes is to teach the specific set of skills of industrial automation with a view to place them in a particular sector.
c. In-house training centres of industries
The main aim here is to onboard candidates with skills required for the success of the underlying business or technology. Hence, in the context of deploying automation to the end application, a fresh student entering an MSME or the new employee being onboarded does not get a well-rounded training.
d. Methods and assessment:
The traditional way of measuring students’ learning is in the form of in-class quiz, mid semester exams and end semester exams with lab sessions. Alternatively, a more flipped classroom measures the progress of students more frequently where the faculty plays a role of a mentor when needed and imparts conceptual knowledge while the experiments are going on. It can be argued that the traditional structured lab sessions, which follow or run in parallel with the lecture sessions, achieve just the right balance. However, the boundaries between disciplines are blurred in industry. Hence, a traditional lab session may not be sufficient. This is particularly true for courses on Automation.
In summary, a university follows written exams and lab experiments to assess students. A vocational training centre would employ an emulator project and an industry would deploy the trained engineer to a somewhat beginner level project. None of them, however, employ continuous assessment to the degree needed. Thus, there is a strong need to find innovative ways to teach multidisciplinary courses such as industrial automation.
3. Role of AI in automation and teaching automation
MIT published a report on Online Education Policy Initiative in 2016 which recommended building a digital scaffolding that implements the following key ideas:
1. Short duration instruction for better engagement.
2. Recall of previously learned material.
3. Repetition with increasing complexity.
4. Adapting to varying needs of learners.
While existing online initiatives partially achieve one or more of the above four concepts, it is still needed to work more in terms of measuring the performance of students and constantly adapting to their needs. AI works best based on past records of data. It is a good opportunity to apply these capabilities to assist both teachers and students.
The main motivation to bring AI in delivery of modules is the need of reducing onboarding time in industry. It takes anywhere from six months to a couple of years before a fresh recruit becomes ‘billable’. If this time can be reduced using the latest in AI, it would serve two purposes. The organisation will benefit from quick onboarding and the employee will acquire skills faster with lesser human intervention.
4. Proposed Model
AptRaise Technologies Pvt Ltd is a rapidly growing technology organisation that provides solutions encompassing the above methodology. ‘ed up me’ is a unique AI powered framework offered by AptRaise Technologies for faster onboarding of new employees. The proposed model has the following four major steps towards delivery of modules on industrial automation:
1. Short modules based on targeted skills
2. Delivery of concepts on digital mobile technology
3. Reinforcement using portable kits where students can perform experiments
4. Continuous assessment and help using AI
5. A platform where industry and academia participants come together.
Brief summary of each is explained below.
Segments of industrial automation are divided in short modules. This is already being implemented in internal training of several organisations. Design of modules can vary based on applications. For example, a process heavy manufacturing set-up can focus more on basics of sensors, PLC and actuators whereas a warehouse with logistics challenges can lean more on mobile robotics.
Digital delivery on mobiles/tablets
Learning does happen more frequently when the material is available on a mobile instrument. Again, this is being implemented partially by popular apps for language and science courses. This can be implemented for short modules of industrial automation.
Experimental kits and a combined platform
Automation kits are now widely used for a particular industry which wishes to train their current and future engineers. A platform can then be created where real use cases of industry can provide the correct motivation where a trainer can train using a combination of kits and AI based delivery.
One of the drawbacks of prevalent methods of assessment is that the opportunity to revise the material arises only a few times. All existing ways miss out on personalised measurement. It is not possible to keep measuring in real-time what a human mind understands. In an AI based digital platform like ‘ed up me’, material can be presented in short modules and users can access as and when needed.
‘ed up me’ is an AI powered interactive knowledge delivery framework built by AptRaise Technologies Pvt Ltd. This framework implements the model discussed in previous sections. Interested readers can place queries regarding the article to email@example.com and access more information on www.aptraise.com.
Dr Harshal Oza is an academician and one of the co-founders of AptRaise Technologies Pvt Ltd. In the last decade of his experience in academia, Harshal has developed expertise and passion for training young engineers. As a faculty member in a university, he has taught several multidisciplinary courses and developed labs of undergraduate engineering curriculum. He has a B.E in Mechatronics Engineering and Masters and PhD in Control Systems Engineering. Before joining academia, he received scholarships for his PhD and postdoctoral positions from the Engineering and Physical Science Research Council (EPSRC), UK. Harshal’s professional engineering experience ranges from control design in diverse applications in textile and packaging machines to detailed engineering review of instrumentation in petrochemical refining. He has published articles in reputed international journals published by IEEE, SIAM, AIAA and IMA. His papers have also been published in reputed international conferences such as IEEE-ICRA (Robotics and Automation) and IEEE-CDC (Control and Decision). Harshal has secured funding from DST for sponsored research. Dr Oza is an active member of IEEE, IET and ASME. He routinely reviews papers submitted in journals of IEEE, Elsevier and Wiley.