COLLEGE OF COMPUTING AND INFORMATION SCIENCES

Mid-Term Assessment Fall 2020 Semester

Class Id 105191,105271,105159 Course Title Artificial Intelligence
Program BSCS and MCS Campus / Shift Main Campus / Morning
Date 24th – October 2020 Total Marks 30
Duration 02 hours Faculty Name Siraj Munir
Student Id 63758 Student Name Ali Salman Hassan

Instructions:
• Filling out Student-ID and Student-Name on exam header is mandatory.
• Do not remove or change any part of exam header or question paper.
• Write down your answers in given space or at the end of exam paper with proper title “Answer
for Question# _ _”.
• Answers should be formatted correctly (font size, alignment and etc.)
• Handwritten text or image should be on A4 size page with clear visibility of contents.
• Only PDF format is accepted (Student are advise to install necessary software)
• In case of CHEATING, COPIED material or any unfair means would result in negative marking
or ZERO.
 A mandatory recorded viva session will be conducted to ascertain the quality of answer scripts
where deemed necessary.

 Caution: Duration to perform Mid-Term Assessment is 02 hours only. Extra 01 hours are given
to cater all kinds of odds in submission of Answer-sheet. Therefore, if you failed to upload
answer sheet on LMS (in PDF format) within 03 hours limit, you would be considered as
ABSENT/FAILED.
 Understanding the questions is the part of assessment.
 It’s always good to have choices in life but this exam doesn’t has choice 
 All questions are correct if you think there is any mistake so take your assumption and try to solve
it.
 This paper is open book but not open discussion, so you can use book and class lectures for
getting help but copy paste from any source is strictly not allowed if you found doing such stuff
you’ll get 0 marks. To avoid such conditions, use your own wording.
 Restrict your discussions to classroom knowledge.
 You are expected to complete the paper within three hours i.e. 12-3 pm, however to accommodate
electricity or internet failures and your convenience total time is 12pm to 6pm.
 You will get 10% bonus of earned marks, if the answer script is uploaded as PDF file within 3.5
hours of start time i.e. by 3:30 PM on Google classroom.
 You will upload same answer script to Kiet LMS and Google Classroom if any difference found
between both submissions you’ll get 0 marks.
 Deadline to upload answer script to Kiet LMS is 7-8 pm. Before and after allocated time your
submission won’t be accepted.

I have read all notes and I assure that all answer provided in this answer script are purely my contribution.
For getting help I used book and related class material but I didn’t discuss any of the question with my
classmates/ seniors.

Write your name here: _____________

Question 1: ROBOW (3+3+2+3+2+3+2+2+4+2+2 = 30 Marks)

Page 1/ 5
The manufacturer of the ROBOW, ROBOW, is a new, and much smarter vacuum cleaner. The ROBOW
uses different artificial intelligence based models to scan room size, identify obstacles and remember the
most efficient routes for cleaning. The self-deploying ROBOW can also determine how much vacuuming
there is to do based on a room's size, and it needs no human assistance to clean floors. The company
completed its first year as a purely consumer-focused business in 2018, pulling in $9.9 million in revenue,
and has shipped more than 10 million ROBOW since 2000.

A. Define what type of agent is ROBOW. Draw architecture for ROBOW.

ANSWER:

It is a learning agent.

Page 2/ 5
B. Explain in which type of environments ROBOW can work. Justify (use 2-3 sentences)
ANSWER:

It can work in fully observable, deterministic and static environment because it has a room model
(by camera) which can be seen by its actuators and sensors. It also has discrete actions and it is
also deterministic because it has limited number of actions.

C. Define PEAS for ROBOW.

ANSWER:

Performance measure: Electricity cost, Cleaning cost.

Environment: Hospitals, Offices, Homes, Modern day parks.
Actuators: Wheels, Dirt sucking pump.
Sensors: Camera, Sensors.

D. Write pseudocode/algorithm for architecture you defined in part A for ROBOW. Can you evaluate
its worst case complexity in bigO notation?
ANSWER:

Algo:
LEFT, RIGHT, FORWARD, BACK, DO NOTHING, SUCK DIRT.

It`s worst case complexity will be O(n).

E. What will happen if the state space of ROBOW would increase n x n. Does it affect its
performance measure? If yes, then how much? Evaluate it.
ANSWER:

Yes, it will affect performance measure because as the room size will increase, the amount of dirt
will also get increase and it will take more time for searching and cleaning the dirt. If state
space(room dimensions) will be 7x7, it will take 10 minutes to clean and as we will increase the
dimensions more and more, the time for cleaning will also increase.

F. You studied multiple searching approaches throughout this course. Justify which search approach
is suitable for ROBOW. (use 3-4 sentences)
ANSWER:

A* search algorithm will be suitable for this scenario because it works on heuristics cost which
make him better and more efficient to calculate the path by combining path cost and heuristic cost
for calculating the path.

G. Let say you come to a situation where clean room can become dirty and location/ size of room is
dynamic/ unknown. So, will ROBOW use its learning experience in such case? What should it
learn? If not, why not?
ANSWER:

In such a case, ROBOW will be using his routes that he learned to work and during the action, he
will calculate the size and location of a room and will compare it with its model of room (by
camera). So, if this condition occurs in future again, he will recall it from his memory.

H. Show that ROBOW is rational.

ANSWER:

Page 3/ 5
 ROBOW is rational because his percepts have possible action states which helps to make his
performance better and even more efficient by having its own model of room (by camera) and he
knows how much dirt in room to clean.

Question 2: Consider following network:

Devise problem and goal formulation for the network starting from S and ending on G. Next write
traversal for the given network using A* search algorithm and Uniform cost search algorithm.
Which one performed well and why?
Heuristic h(n): S:10, A:5, B:50, C: 10, D: 10, G:0
ANSWER:

A* search algorithm will perform well according to given scenario because it is providing best path
route by using path cost as well as heuristic cost to give us exact data as compared to the uniform
cost search algorithm.

Page 4/ 5
Page 5/ 5