Professional Documents
Culture Documents
Imperial College of Engineering: and Research
Imperial College of Engineering: and Research
Submitted By
Mr. Rahul M. Shelar
Mr. Parvez Z. Sayyad
Mr. Ankush S. Patil
Mr. Pranav S. Watile
CERTIFICATE
This is to certify that Mr. Rahul Shelar, Mr. Parvez Sayyad, Mr.
Ankush Patil & Mr. Pranav Watile has successfully completed
the Project Stage-I entitled “Automation of industrial
articles with using bin piking by lab view ” under my
supervision, in the partial fulfillment of Bachelor of
Engineering – Mechanical Engineering, by Savitribai Phule
Pune University.
Date: / /2021
Place: Wagholi
Seal
ACKNOWLEDGEMENT
It’s my great pleasure to present project report entitled “ Automation of industrial articles
with using bin piking by lab view ”. This projects an outcome of various efforts by us in
Collating and identifying the sources of information and knowledge. I use this occasion to thank
my guide with whose guidance this effort would not have borne fruits. I find no words to express
my gratitude to Prof. Nilesh Patil. Who not only advised and guided me during the report
writing but also answered all my queries concerning collection of data, proper structuring of the
report, and its improvement. I would also thankful of my friends which gave me a proper guide
line about my subject. I am virtually indebted to the Head of the Department Prof. N. S. Biradar
from Imperial College of Engineering and Research who have bestowed all their blessings in the
form of guidance which was the leading light to complete this seminar report. I would also like
to thank our principal, Dr. R. S. Deshpande who provided me valuable support in completion of
seminar by providing me different facilities in college and by giving permission for working out
of college.
Bin Picking usually involves numerous component variants that require effective
sorting of components. Here, picking activities are often performed manually, but the
prospect of robotics for bin picking has potential to improve quality while reducing
man-hour consumption. Robots can make use of vision systems to learn how to perform
their tasks. This project aims to understand the differences in two learning approaches,
supervised learning, and unsupervised learning. The proposed work investigates the
benefits of combining a vision system and deep learning policies with a collaborative
robot for sorting purpose in industries. Finally, the project analyzes the component and
then sort the components with high performance accurately.
INDEX
1. Introduction
2. Literature Review
3. Methodology
4. Advantages
5. Limitations
6. Applications
7. Reference
INTRODUCTION
Bin Picking usually involves. numerous component variants that require effective
sorting of components. Picking and sorting activities are often performed in warehouses
or at logistics workspaces, e.g. as kit preparation, in order to arrange materials for the
assembly processes, and high levels of quality, flexibility, and productivity is essential.
Picking processes involve extensive material handling work, but it also involves
considerable engineering work with keeping the process up to date with changes in the
product structure and SKU (Stock-Keeping Unit) assortment. The engineering work is
critical for normal operation of the picking process, and it is important that the engineers
are equipped with effective tools to this end.
Using robotics for picking and sorting activities has the potential to improve
quality and productivity. While robots tend to perform effectively in deterministic
settings. where all critical factors are known before task execution, the application of
robots to picking activities usually implies dealing with materials arranged as received
from suppliers, in form of picking and sorting of randomly organized materials inside
bins - often referred to as bin picking. Bin picking is a challenge in most robotic
applications, owing to the precise information needed about the items position and
orientation for the robot to effectively perform its task.
Vision systems referring to a scanner for taking 2D or 3D images of objects, and a
software for analyzing the images to extract useful information can be an effective
support for robotic bin picking applications, whereby robots can learn to recognize
components in order to perform their tasks. Robots can learn in accordance with two
principal approaches: supervised and unsupervised. With the former, at person shows
the robot what to recognize, for example by means of indicating the shapes of items to
pick on a set of images, a process hereon forth denoted as annotation. With the latter, the
learning takes place automatically by means of a deep learning model which is capable
of distinguishing objects and object features, ideally without the support of a person.
Emde, S., Boysen, N.:- Mixed-model assembly usually involves numerous component
variants that require effective materials supply. Here, picking activities are often
performed manually, but the prospect of robotics for bin picking has potential to
improve quality while reducing man-hour consumption. Robots can make use of vision
systems to learn how to perform their tasks. This aims to understand the differences in
two learning approaches, supervised learning, and unsupervised learning. An experiment
containing engineering preparation time (EPT) and recognition quality (RQ) is
performed. The findings show an improved RQ but longer EPT with a supervised
compared to an unsupervised approach.
Simonetto, M:- compare the engineering preparation time and the recognition accuracy
of two learning approaches: supervised and unsupervised. The two approaches were
compered in terms of five different categories and criteria with respect to the ability of
the vision system to provide a robot manipulator with reliable information in order to
carry out bin picking of components in an application for kit preparation.
M. Alonso:- This paper presents the details of a collaborative robot cell assembled with
off-the-shelf components designed for random bin-picking and robotic assembly
applications. The proposed work investigates the benefits of combining an advanced
RGB-D vision system and deep learning policies with a collaborative robot for the
assembly of a mobile phone. An optimized version of YOLO is used to detect the
arbitrarily placed components of the mobile phone on the working space. In order to
overcome the challenges of grasping the various components of the mobile phone, a
multi-gripper switching strategy is implemented using suction and multiple fingertips.
Finally, the preliminary experiments performed with the proposed robot cell
demonstrate that the increased learning capabilities of the robot achieve high
performance in identifying the respective components of the mobile phone, grasping
them accurately and performing the final assembly successfully.
METHODOLOGY
We started working on project with some literature review, need or to gather all
information related with this project. Find the type, design and the system used on the
development of machine, also including the differences for each stage in the market. All
the information was collected from internet, journal, reference book and people. We
learnt that there are Robots in the markets which can sort things manually and needed a
professional worker. So, we decided to make a Bin picking robot which can detect the
components and sort them using a programming software.
The main purpose of the robot is to classify the components and then transfer them to
their respective bins. So, there is a camera fitted on the main grabbing am, which grabs
the component using pneumatic system and the camera analyses the component and
shows on the screen, which allows the robot to understand the component and sort it. So,
the robot works in 3 DOF which allows the robotic arms to move in their maximum
This Bin picking Robot allows to increase the production capacity with almost no
errors. It will reduce the human efforts and will only require an observer to check if the
robotic is sorting properly of not which can be done by a low skilled worker. It can be
used in Food industries and other similar industries. We are using a programming
software to input all the command to the robot. The software also uses the camera on the
robot arm and does the calculations according to the software commands. The software
we use is LabVIEW which allows to control all the robotic movements.
8GB Ram
200 GB Storage
i3 Processor
ADVANTAGES
Mass production
Higher quality
Increased efficiency
Improved working environment
Increased profitability
Reduce longer working hour
No skilled labour required
Less cycle time
LIMITATIONS
Programming is Difficult
It is slow working but very precise
APPLICATION
Welding, quality inspection, welding & material handling, bin picking application
REFERENCES
[1] Festo cyper-physical (cp) factory modules (accessed 22-01-2020). URL www.festo-
didactic.com/int-en/learning-systems/ learning-factories,cim-fms-systems/cp-factory/
[5] M. Alonso, A. Izaguirre, M. Grana, ˜ Current research trends in robot grasping and
bin picking, in: The 13th Int. Conf. on Soft Computing Models in Industrial and
Environmental Applications, Springer, 2018, pp. 367–376 (2018).