Assignment 2 Artificial Intelligent

ASSIGNMENT 2
ARTIFICIAL INTELLIGENT
SUBMITTED TO SUBMITTED BY
MR PRABIN KU.ROUT LIZA SETHY(BFT/17/1139)
BHASWATI PANDA(BFT/17/415)
INTRODUCTION :ABOUT TRASHBOT
 Garbage is a global problem that affects all

living beings.
 Trashbot; invented by CLEAN ROBOTICS in
15 Nov 2016, to resolve such problem.
 World’s first “smart” waste bin; automatically
sorts recyclables from everyday trash.
 It’s a combination of next-generation robotics,
advanced sensor technology, and the AI that
helps in-
 Categorizing unwanted materials.
 In auditing waste generation.
 And wirelessly updating itself.
PROBLEMS
 Overflowing garbage bins/waste causes soil and water pollution.
 Bin labels are not being followed correctly.
 The open containers are proving to be a breeding place for germs;
causing diseases.
 Direct handling of overflowing waste exposes for health risks
 The traffic congestion due to slow moving vehicles like trash pickup
truck.
 Hence, huge sum of money is wasted.
PARTS
 single opening for all trash
 open chamber
 cameras and sensors
 Three dumping bins
 mechanical flap system- single recycle
chamber or multi recycle chamber
WORKING
 Garbage; thrown into the TrashBot with single opening.
 Cameras and sensors quickly analyse the type of discarded item.
 It categorizes the data into; type of garbage
quantity of each type of garbage
respective waste disposal method
 The item is weighed on a Teflon-coated plastic shelf; drains the excess
liquid
 Clean Robotics software will decided whether the item is destined for the
landfill or recycling facility
 IOT based trashcan will automatically and continuously update the status
of the trashcan to the Municipality.
 The municipality workers get the notification and they can remove the
garbage.
PROCESS FLOW
IMAGE PROCESSING IN TRASHBOT
 Image of the trash is captured by the RBG camera
 Once the food is recognized, it is compared with the database. A database
consists of pre-loaded nutrition information about various food items.
Therefore along with the weight of the food and the type of the food,
 Sent to the processor.
 Processor transmits the image to a unit which is trained using an
algorithm.
 The processing unit identifies the image of the garbage, classifies it into its
respective category and transmits the data back to the processor.
 Processor then guides the mechanical device.
PROCESS FLOW
INPUT IMAGE FROM CAMERA
SEGMENTATION
FEATURE
EXTRACTION COLOUR TEXTURE
SIZE FEATURE
FEATURE FEATURE
CLASSIFICATION
INPUT
To distinguish between different kinds of inorganic waste.
For Eg: program to identify a plastic bottle thrown into trashbot.
1.SEGMENTATION
 GREY SCALE CONVERTION :RGB image in the trashcanis converted
to grayscale by OpenCV,
 Then BINARIZATION
I. This stage removes the unnecessary features from the image.
II. The image of the waste is represented by means of the area of
the object.
RGB image grayscale image binarized image

FEATURE EXTRACTION
 The objective of feature extraction process is to represent the raw image

in its reduced and compact form
 It allows the CNN to detect features/objects even if it does not look
exactly like the images in it's training period.
 Even if the same waste is turned, if it is not centered, or its size is
different or defected.
FEATURE EXTRACTION
 Image preprocessing: image resizing,
filtering, getting silhouette image and region
properties measurement for the bottle’s
image.
 Bounding box image algorithm:

I. Minimisation of background and maximisation
of object is done.
II. histogram of intensity feature values was
extracted from the silhouette of grey level
image
 Then the object is segmented and the

region of interest is further analysed.
3. CLASSIFICATION ALGORITHM:
 Evaluates compares each piece of trash

with its knowledge base.
 k-NN (K Nearest NEIGHBOUR) classifier;

classifies a new item by searching for its k
nearest trained items.
 Distinguish the bottle from other objects
Classification of a defective
bottle
OUTPUT
 After the classification, TrashBot scans and tells the robotic
arm where to dump it using various sensors.(of the three
bins)
 It tells the difference between recyclables and landfill
waste.
TECHNIQUES USED IN THE SYSTEM
 In top of the garbage can the infrared sensor -sense the waste.
 Signal send to the microcontroller - check the waste is degradable
or non degradable.
 A plate (Using DC motor)-open and close according to separation of
the waste.
 Signal controller module connected internet connected
cloud server.
 The message is then sent to the mobile application of the
Municipality nearby or the worker of the particular area workers.
TYPES OF SENSOR USED
1.OBSTACLE SENSOR SYSTEM-

 Location- Above the lid.
 The lid can’t be open mistakenly when walking
nearby.
 Once the lid is open, the Obstacle Sensor system
knows if anyone is still in the sensor zone, so the
lid won’t close before the action is finished.
 Sensor system ensure lid open/close time is right.
2. INFRARED SENSOR :It does infrared imaging

to easily identify if the product is metal or non metal
and degradable or not.
3. DEPTH SENSOR: It gives the 3D image of the

product.
2.WEIGHT SENSOR SYSTEM-
 Helps in keeping tracks- (recyclable material
and landfill waste on a daily/weekly).
 It’s a basis for accurate waste auditing.
 Weighs the amount of food wasted
3.SELF DRIVING TECHNOLOGY

 TrashBot’s mode of transportation-movement
system.
 It uses differential steering system
 four geared motors with one on each wheel.
 motors top speed-3.89 km/h.
 ability to tackle all types of terrain.
 it helps the trashbot to drive itself to fill the
municipal vans.
FEATURES
 It weighs the food and calculates the money wasted.

 It separates the material with accurate precision; 81% sorting rate
 The robot swallows garbage at the rate of around three seconds
per item.
 Continuously learns to recognize new types of trash.
 Gives information about its disposal system.
 They also provide notifications when the receptacles are full,
helping to improve collection efficiencies.
ADVANTAGES
 It performs intelligent compaction of waste by monitoring fill level in
real-time using sensors.
 Dynamic routing
 Infrastructure and cost reduction
 reduces manpower requirements to handle the garbage collection
process.
 Helps in CO2 reduction & improve cleanliness.
 Easy to install and maintain.
WHERE IT CAN BE USED?
Due to the great capacity, TrashBot can be deployed at :
 airports, airplanes, malls, stadiums, universities, office buildings that
handle large quantities of garbage.
 highly populated urban areas, for eg Japan; heavily investing in efficient
garbage disposal mechanisms.
:airport
DISADVANTAGE OF TRASHBOT
 Trashbot cannot detect the level of garbage. Whether it can
take more load .
 Trashbot is only limited to be used in land not in water.
 Trashbot doesnot support locomotive moment. Hence it cannot
travel by its own.
 It doesnot have odour control mechanism.
SOLUTION
 DETECT THE LEVEL OF GARBAGE
 New system -ultrasonic sensor can be used as input and placed at
the maximum level of the garbage bin.
 ultrasonic sensor- Garbage level measurement.
 ARM microcontroller- controls system operation whereas
everything will be connected to a website.
 At the same time, the level of garbage also will display on LCD to
allow user to know the level of garbage in the dustbin without open
it.
 The four ultrasonic sensors can be connected to ARM
microcontroller to detect the level of garbage of each bin based on
the depth of the bin.
 At the same, these four ultrasonic sensors can connect to a wifi
module to make sure the data transfer and display on website.
 The LCD should interfaces with ARM microcontroller which will
display the percentage of the garbage for each bins.
 In this work, the system will try to monitor the depth of the garbage
based on garbage type.
 ODOUR CONTROL MECHANISM
 Dryer sheets-The fresh-smelling sheets can avoid the nasty
trash can odors.
 Once the trashbot is emptied, dryer sheets can be placed
underneath the trashbot. The sheets will absorb bad odors and
leave a pleasant scent behind.
 There can be a change in the sheet each time you take out the
trash.
 Charcoal- Charcoal is proven to filter out odors.
 In this case we don't need activated charcoal; AI should be
used in a manner so that it can get infused underneath.
LOCOMOTION MECHANISM
 The trashbot can travel in the predetermined path; by combination of motors,
drivers, and sensors connected to the Arduino.
 The ultrasonic sensors will act as input to the Arduino.
 Two motors are mounted on the two sides of the shaft
 Connected to Arduino to perform rotating mechanism
 Aim is to collect garbage which is of similar dimensions to that of juice
cartons, plastic bottles, crushed papers,
all light items whose height is between 5 to 20cms.
 When the sensor detects static obstacle, the shaft will rotate and the garbage
is pushed into a corresponding bin, by the mechanical arms.
Servo motor
Bins
Arms
Wheels Trashbot
MECHANISM
FLOW DIAGRAM
FLOATING TRASHBOT FOR WATER
 Trash; make the waterbody unsafe for use.
 It can be like a motor boat; one can drive it with a web connection
 The program will direct TrashBot to pick up trash in its immediate vicinity
and move to the collection point
 It has Motor operated wheel to run. It has four DC Motor of 12V, 7.6
Ampere.
 It will suck all the waste from the river and sort it out.
 a powerful Wi-Fi transmitter, to keep a track of it
 Or can be remote controlled, through RF transmitter and receiver
Motors
Collecting arms
bins
Conveyor belt
REFERENCE
 https://medium.com/syncedreview/trashbots-
boost-recycling-at-pittsburgh-airport-
258004877bcd
 http://www.cleanrobotics.com/2016/11/15/Introduci
ngTrashbot/
 https://medium.com/makersbootcamp/sort-it-out-a-
smart-trash-bin-from-cleanrobotics-66a46bedff11
THANK YOU

Assignment 2 Artificial Intelligent

Uploaded by

Copyright:

Available Formats

Assignment 2 Artificial Intelligent

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Assignment 2 Artificial Intelligent

Uploaded by

Copyright:

Available Formats

ASSIGNMENT 2

 Garbage is a global problem that affects all

INPUT IMAGE FROM CAMERA

RGB image grayscale image binarized image

 The objective of feature extraction process is to represent the raw image

 Bounding box image algorithm:

 Then the object is segmented and the

 Evaluates compares each piece of trash

 k-NN (K Nearest NEIGHBOUR) classifier;

1.OBSTACLE SENSOR SYSTEM-

2. INFRARED SENSOR :It does infrared imaging

3. DEPTH SENSOR: It gives the 3D image of the

3.SELF DRIVING TECHNOLOGY

 It weighs the food and calculates the money wasted.

You might also like