Roll Number:__________________________
Thapar Institute of Engineering & Technology, Patiala
Department of Mechanical Engineering
End Semester Test (EST), December 2023
B E-Mechanical Engineering (VII Semester) Course Code: UME726
Course Name: IoT Enabled Automation
and Machine Learning
December 04, 2023 Monday, 14:00 – 17:00 Hours
Time: 3 Hours, M. Marks: 70, Weightage: 35% Name of Faculty: Dr. Anant K Singh
Note: Attempt any 7 questions
Assume missing data, if any, suitably
Marks CO BL
Q1. What key goals does machine learning aim to achieve? What challenges does (10) -- L1
machine learning face for its implementation in the manufacturing industry? Write
your answer in five points for each part of this question.
Q2. Differentiate between supervised and unsupervised learning with the help of a (10) CO2 L2
practical example and sketches.
Q3. Why are training data and testing data required in machine learning? How do you (10) CO2 L3
divide the training and testing for the given sensor's large dataset using holdout
validation? Write the Matlab codes for dividing 80% of training data and 20%
testing data for any given dataset used in the manufacturing industry for
classification algorithms.
Q4. (a) How to declare and call functions in Python programs? Illustrate with an (10) CO2 L3
example script. (b) List and explain five most commonly used built-in functions in
python.
Q5. (a) Why are NumPy, Pandas, Matplotlib, Seaborn, and SciKit Learn used in Python (10) CO2 L3
machine learning? (b) Write the important Python code in the various steps required for
implementing “Regression SciKit Learn” Python machine learning for any practical
application where Python machine learning is most required.
Q6. How can the remaining useful life of any machine be predicted using machine (10)
learning predictive maintenance in Matlab? Explain the purpose of different CO2 L2
machine learning estimator models used in predicting the remaining useful life of
the machine.
Q7. What is the role of the confusion matrix in implementing the classification machine
learning algorithm? Analyze a confusion matrix for any three activities used in the (10) CO3 L4
manufacturing or automotive industry and analyze its model performance for the
classification algorithm.
1/2
Q8. The M2M separating station required for implementing IIoT in modern manufacturing industry is (10) CO1 L3
shown in Fig.1. The separating station consists of 2 Conveyor modules. The station detects the
height of incoming workpieces and allocates them. A diffuse sensor at the start of the conveyor
detects the inserted workpiece. A distance sensor above the stopper detects the inserted
workpiece and controls the further material flow. Depending on the rating, a separator is
switched. If the pick-up point and the conveyor section is not occupied, the stopper is retracted
and a workpiece is transported to one of both conveyor belts.
A typical sequence for the separating station is given as follows.
Initial position: stopper advanced, separator retracted and conveyor motors off.
Sequence
1.The conveyor belt motor 1 switches on if a workpiece is detected. The workpiece is transported
to the stopper.
2.The conveyor belt motor 1 is switched off, if the workpiece is detected by the diffuse sensor in
front the stopper.
3.The differentiation of the workpiece takes place by the distance sensor above the stopper.
When workpiece “cylinder body“ is detected, transport to pick-up point conveyor belt 1
4.The stopper is retracted and the conveyor belt motor 1 is switched on, if the pick-up point is
empty (indicator light “workpiece at pick-up point” is switched off). The workpiece “cylinder
body“ is transported to the pick-up point at the end of conveyor belt 1.
5.The conveyor belt motor 1 is switched off, the stopper is advanced and the indicator light
“workpiece at pick-up point” is switched on, when the workpiece “cylinder body” has reached the
pick-up point.
When workpiece “housing“ is detected, transport to conveyor belt 2
6.The stopper is retracted, the separator is advanced and both conveyor belt motors are switched
on, if the conveyor belt 2 is empty (indicator light “buffer full” is switched off). The workpiece
“housing“ is transported to the end of conveyor belt 2.
7.The stopper is advanced, the separator is retracted and the conveyor belt motor 1 is switched
off, if the workpiece “housing” is detected by the through-beam sensor at conveyor belt 2
(indicator light “buffer full” is switched on).
8.The conveyor belt motor 2 is switched off and the indicator light “buffer full” is switched off,
when the workpiece ”housing” has reached the end of conveyor belt 2.
Conveyor belt-2 Pin allocation: %IX1.0 (Start
Conveyor belt-1 pushbutton), I0 (Workpiece at beginning
Stopper from bottom
of conveyor 1), I1 (Workpiece in middle
Separator of conveyor 1), I2 (No workpiece at end of
conveyor 1), I3 (Distance sensor), I4
Workpiece cylinder type (Workpiece at beginning of conveyor 2),
Workpiece housing type I5 (Workpiece at end of conveyor 2), Q0
(Conveyor 1 forward), Q1 (Conveyor 1
reverse), Q2 (Advance feed separator), Q3
(Activate stopper), Q4 (Conveyor 2
forward), Q5 (Conveyor 2 reverse).
Fig.1 The separating station.
To control these automatic sequences of the above operations, as shown in Fig. 1,
(a) write the code to control and visualize the given sequence of operations in the soft PLC
Ladder Logic using the CODESYS Driver and OPC-UA Protocol for communication.
(b) Explain briefly the working principle of the sensors used in the above application.
(c) How is IIoT to be implemented for the given M2M separating station? Show the connectivity
with IIoT devices and protocol.
2/2
