Course Code : 19UCSPEX01
Course Name : FOUNDATIONS OF AI
Semester/Year : B.E (CSE) VI SEMESTER/ III YEAR
Subject In charge : Martina Jose Mary.M, AP/ CSE
Q.No Unit No. Assignment Task/Questions Marks CO PO
Objective:
• Explore the basic Artificial Intelligence approaches.
• Acquire knowledge on problem solving agents and search algorithm.
• Demonstrate how to represent knowledge in solving Artificial Intelligence task.
• Introduce the different ways of designing software agents
• Explain the various applications of Artificial Intelligence.
1 UNIT 1 Current AI systems often specialize in one 16 CO 1 PO 1,
type of task (e.g., rule-based, learning-based, PO 2,
goal-driven). Design a hybrid intelligent agent PO 3,
that combines multiple AI characteristics PSO 1,
(reactive, goal-based, utility-based, learning). PSO 2
Describe its architecture, decision-making
flow, and how it would handle an AI-specific
problem such as real-time medical diagnosis or
stock market prediction.
2 UNIT 2 Create a real-world problem that can be 16 CO 2 PO 1,
formulated as a CSP, ensuring it involves PO 2,
complex constraints (e.g., scheduling under PO 3,
dynamic conditions, multi-robot task allocation, PO 4,
or resource optimization in smart grids). Model PO 9,
the problem mathematically, propose different PO 12,
constraint-solving techniques, and compare PSO 1,
how various search strategies perform in PSO 2,
solving it. PSO 3
3 UNIT 3 An autonomous vehicle approaches an 16 CO 3 PO 1,
intersection where the traffic light is green for PO 2,
vehicles. However, a pedestrian unexpectedly PO 3,
starts crossing. The AI system, which relies on an PO 4,
ontology for categorizing objects (e.g., PO 5,
pedestrians, traffic lights) and events (e.g., PO 8,
crossing, stopping), must reason using default PO 12,
knowledge: PSO 1,
Default Rule: "Pedestrians usually wait when the PSO 2,
vehicle light is green." PSO 3
Exception Handling: "A pedestrian crossing on
green may indicate an emergency or rule-
breaking behavior."
Given that the system must infer mental objects
(e.g., the pedestrian’s intent) and update its
reasoning dynamically, how should it adjust its
decision-making process to balance safety and
efficiency while refining its ontological
categories in real time?
4 UNIT 4 Development of a smart city's autonomous 16 CO 4 PO 1,
delivery drone fleet, your task is to address the PO 2,
technical and ethical challenges associated with PO 3,
PO 4,
this innovative technology. Your case study
PO 5,
should provide comprehensive solutions to PO 8,
ensure seamless coordination, efficient PO 12,
navigation, and safe operations. PSO 1,
Address the following aspects of the PSO 2,
autonomous delivery drone fleet: PSO 3
• Recommend a suitable communication
protocol (e.g., MQTT, CoAP,
LWM2M) to ensure seamless
coordination between drones and the
control center. Justify your choice based
on factors like latency, scalability, and
security.
• Propose a message prioritization
framework to handle multiple urgent
messages (e.g., low battery alert vs. no-
fly zone update). Explain how your
framework ensures timely and effective
decision-making.
• Select a suitable Machine Learning
technique (e.g., Reinforcement
Learning, Neural Networks) for real-
time route optimization. Explain how
your chosen technique adapts to
dynamic environments and optimizes
delivery routes.
• Design an algorithm or decision-making
framework for modifying a drone's
flight path when encountering an
unexpected obstacle. Consider factors
like sensor data, environmental
conditions, and drone dynamics.
• Discuss how the system should handle
privacy concerns related to drone
surveillance in urban environments.
Propose measures to ensure minimal
data collection and secure storage.
• Outline the steps to be taken to ensure
safe drone operations during
emergencies like medical evacuations or
firefighting. Consider factors like
priority routing, communication
protocols, and drone deployment
strategies.
5 UNIT 5 In an AI-powered assistive communication 16 CO 5 PO 1,
system for individuals who are deaf and dumb, PO 2,
how can Natural Language Processing (NLP), PO 3,
PO 4,
Machine Translation, and Speech Recognition
PO 8,
be utilized to enable seamless interaction with PO 12,
healthcare providers, emergency responders, or PSO 1,
everyday services? Provide a real-time scenario PSO 2,
where the system interprets sign language, PSO 3
converts text to speech, or translates gestures
into multiple languages to enhance accessibility
and inclusivity.
