VIT

VIT
Vellore Institute of Technology
Reg. No. : 2|8LC22.
FinalAssessment Test(FAT) -Nov/Dee 2024
Programme B.Tech. Semester Fall Semnester 2024-25
Course Code BCSE429L Faculty Name Prof. Parvathy AK
Course Title Cyber Physical Systems Design Slot C1
Class Nbr CH2024250101281
Time 3 hours Max. Marks 100
General Instructions
" Write only Register Number in the Question Paper where space is provided (right-side at the top) & do
not write any other details.

Course Outcomes
1. To understand, analyze and design the communication components, computation models and coordination
models of a CPS
2. To evaluate the design of CPS system through verification and validation
3. To design coordination models for heterogeneous CPS
4. To state the challenges and the opportunities of CPS

Section -I
Answer all Questions (10 x 10 Marks) *M- Marks

Q.No Question *M CO BL

01. As an engineer, you are instructed to evaluate the suitability of three hardware platfoms 10 1 3
RISC-based microcontroller, x86 processor, and FPGA for various applications. Analyze the
power profiles, performance characteristics, and suitability of each platform for Autonomous
Drone Navigation System, Weather Forecasting Supercomputer, Smartphone Application
Processor, Home Automation Controller, Industrial Robot Control System. Give justification for
each.

02. a) Discuss how the presence of cache memory influences the overall perfonance of mobile 10 3
applications, especially in tasks requiring real-time processing. [3]

b) Given the output of an A/D converter is 11010000 with a reference voltage of 12 Volts,
calculate the actual input voltage. If the actual input voltage measured is 8 Volts, determine the
error rate. [4]

c) Discuss whether increasing the clock signals will improve the accuracy of the ADC and
suggest methods to enhance the output of the ADC. [3]

Page 1of 4
 2 2
management system that uses a fluman-in-the- 10
05. Designan architecture for asnart home energy consumption
The system mnonitors encrgy
Loop (HITL) predictive model-based control system. while
in real-time to optimize energy usage
patterns, lemperature, and humidity levels sensors,
model uses data from occupancy
maintaining comfort for residents. The predictive
settings for lighting, heating, and cooling.
appliance usage, and weather forecasts to adjust specify the
sensors, predictive modcls, control units) and
a) ldentify the key components (c.g.,
role of each in the system. (3)
users to
b) Explain how the system would
incorporate Human-in-the-Loop elements, allowing
controls based on personal preferences or comfort levels.(4]
adjust or overide automatic
would improve model predictions over time, ensuring
c) Describe how feedback from residents
eflectively.(3]
the system adapts to their behavior andpreferences
10 3 4
processing system that receives an input stream of
04. Design an SDF model for areal-time signal
normalizes each amplitude value to fit within a
sound amplitude values from a microphone,
smoothing filter over a sliding window of the last
specilic range, and then applies an average
five samples. In your design:
normalization, smoothing filter) and describe their
a) Identify the key actors involved (e.g.,
roles in processing the input sequence. [4]

these actors, ensuring that each actor's output is

b)Define the data flow connections between
correctly timed and synchronized with the next step. [3]
there are any, such as for the sliding window in
c) Specify the delay or buffering requirements if
the smoothing filter, to ensure accurate output. [3]
10 3 4
modeled as actors in an SDF system:
05. Consider a video streaming system with four components
producing 6 tokens per cycle (each
Frame Grabber: Captures frames from the video source,
token represents a frame).
compresses the frames, and produces 4
Encoder: Consumes 6 tokens from the Frame Grabber,
tokens per cycle (representing encoded frames). for transmission,
prepares them
Network Transmitter: Consumes 4 tokens from the Encoder,
and produces 2 tokens per cycle.
the frames for display,
Decoder: Consumes 2 tokens from the Network Transmitter and decodes
consuming 2 tokens per cycle.

actor's token production and

a) Derive the Balance Equations for the system, showing how cach
SDF model. [31
consumption rates must balance for a consistent
of each actor to ensure
b) Design a Consistent SDF Model by adjusting the repetition counts
or starvation.[4]
each actor operates without buffer overflows

c)Provide an Example of an Inconsistent SDF Model, explaining why it cannot be mnodeled

consistently and identifying the resulting issues with token low or buffer capacity in the video
streaning system. [3)

Page 2 of4
 10 3 4
rcquiring critical care are placed
hospital's nconatal intensive care unit (NICU), newborns
In a detcction test is
(06,
monitoring. For those at high risk of infcction, an carly
under constant treatments:
results, infants may undergo different
performed every two hours. Based on tlhe test
infant
Stable Condition: If test results indicate stability (i.c., normal vital signs), the
a)
intervention.
continues regular monitoring without additional
infection are detectcd, the infant is given an
b) Infection Risk Detected: If carly signs of
antibiotic injection.
infection is confirmed, the infant is moved to an
c) Severe Infection Detccted: If a severe
isolation room and given specialized treatment.
automata coordination model for this
Design a Finite State Machine (FSM) model and a timed
to fill in data gaps
NICUinfection detection and treatment process. You may make assumptions
as nccded.
10 3 4
running in the same
07. Consider a simplified railway system with two trains, Train A and Train B,
shared track segment
direction on a single track. There's a signal (Sl) that controls access to a
pass through SI
betwecn two stations, Station l and Station 2. Train A and Train B both nced to
The green light at
to enter the shared scgment. Train Aapproaches SI first, followed by Train B.
the train to wait.
Sl allows a train to enter the shared segment, while the red light requires
Time Constraints:
When a train arrives at SI and the signal is red, it waits up to 5 minutes.
The signal stays green for 3 minutes once it is switched on.
Train A takes 4 minutes to cross the segment, while Train B takes 6 minutes.
Design a timed automata coordination model with justification.
A smart agriculture company is deploying a network of loT-enabled swarm sensors across a 10 4 4
08.
large farm to monitor soil health, crop growth, moisture levels, and pest presence. These sensors
are programmed to gather real-time data, communicate with one another, and adapt their
measurements based on environmental changes. The company's goals are to Optimize Water
Usage, Detect and Control Pest Outbreaks Early and Monitor Crop Health. As the loT specialist
overseeing this project:

a) Identify the types of data that swarn sensors would need to collect to meet the company's
objectives. [2]

b) Analyze the potential challenges in data synchronization and processing across a

decentralized sensor network. How would Swam OS help manage these challenges effectively?
[3]

) Propose an architecture that could handle this data in real time, considering both the need for
rapid decision-making in the field and longer-term data storage and analysis for predictive
insights. (3]

d) Discus any privacy or security nmeasures that should be implemented to ensure the integrity
of the data and prevent unauthorized access.(2]

Page3 of4
09. Ih an ettort to promote sustainable cnergy and minimize power outages, the Government of 10 4 4
India has launched apilot projcct in collaboration witlh aprivate company to analyze the energy
denandand supply patterns across various metropolitan areas. The aim is lo optimize energy
distribution, minimize wastage, and prepare for high- denmand periods. This pilol project is
focused on implementing a smart energy grid that utilizes loT-enabled sensors to collect data on
power consumption, grid performance, renewable cnergy inputs (like solar and wind), and
potential power leakages.
a) Identify the critical data types and sources that need to be monitored in a smart energy grid
[2]

b)Discuss the primary challenges associated with handing large amounts of data from various
sensors across the grid [4]

c) Propose an optimal solution to manage and analyze the collected data efliciently, including
any technology stacks, data processing techniques,or architecture that wouldbest support real
time decision-making and predictive maintenance in a smart grid environment. (4)
10. Acompany that manufactures smart HVAC (Heating, Ventilation, and Air Conditioning) 10 4 4

systems for large commercial buildings is adopting Industry 4.0 technologies to enhance
operational efficicncy, predictive maintenance, and customer satisfaction. Each HVAC system is
equipped with loT sensors that collect data, such as temperature, airllow rates, energy
consumption, and indoor air quality. The company also receives data from a network of
installation partners, maintenance providers, and customers, covering usage pattens, service
history, and location-specific environmental data.

a) Identify the various sources and types of data that the company might be gathering. (3 marks)

b) Discuss the primary challenges in handling this large-scale industrial data. (3 marks)

) Propose an optimal solution for managing and analyzing this data, including any technology
or architecture that would best support this objective. (4 marks)

BL-Bloom's Taxonomy Levels - (1.Remembering, 2.Understanding, 3.Appling, 4.Analysing, 5.Evaluating,

6.Creating)

Page 4 of 4