Professional Portfolio

Professional Portfolio

Kumaresh Kasi ViswanathanMaster of Science -Industrial & Systems Engineering

1

About me

EDUCATION

• Nitte Meenakshi Institute of Technology (May 2011)

– BS in Mechanical Engineering

• Texas A&M University (Aug 2015)

– MS in Industrial and Systems Engineering

2

WORK EXPERIENCE

• National Aerospace Laboratories (June 2010)

– Student Intern

• Tech Mahindra (May 2012)

– Developer

• Ethos Energy (June 2010)

– Engineering Intern

NATIONAL AEROSPACE LABORATORIES

3

Student Intern

• Study and test of Wankle engine to optimize performance, increase fuel economy and power.

• Study of set up of different test rigs used by the Propulsion division to carry out various test

on the aircraft engine.

• Study of MAVs (Micro Air Vehicles) construction for military and civil purposes.

• Basics of CFD (Computational Fluid Dynamics).

• Introduction to Testing of Bearings.

TECH MAHINDRA

4

Junior Sales Representative – Hi – Tech

• Shortlisted prospective clients and prepared business cases as a junior sales representative.

• Interacted with clients through telephonic calls and emails about department’s capabilities.

• Scheduled meetings with prospective clients.

Design Engineer – Dassault Falcon Jet

• Interacted with onsite team on a regular basis.

• Assisted team with 3D design and specialized in 2D drafting for yet another leading aircraft company.

• Applied Six Sigma principles to reduce cycle time and effectively completed project deliverables on time.

Process Engineer – Bombardier

• Administered the optimal process for creating an Aircraft detail part for a leading aircraft company as a Liaised

between the Design team and Onsite Shop Floor team.

• Generated Tool Orders for parts using MAC PAC and modified designs & graphics using CATIA.

ETHOS ENERGY GROUP

5

Engineering Intern

• Contributed to the development of a front end structure of a mobile application called Data

Collect.

• Eliminated the usage of physical records on the shop floor and which helps to reduce

operating time.

• Created Process Flow Maps for the Supply Chain department.

• Interacted with other local and international departments to present more clarity by defining

team roles.

• In-depth understanding of the process flow for Parts Sales and Component Returns.

• Gained practical knowledge of Contract Negotiations, Operations and Supply Chain.

Texas A&M University

• Survey of Optimization - ISEN 620• Theory of Statistics - STAT 610• Systems Engineering Methods and Frameworks - ISEN 641• Probability of Engineering - ISEN 609• Management of Engineering Systems - ISEN 689• Accounting Concepts and Procedures - ACCT 640• Directed Studies - ISEN 689• Simulation Methods and Applications - ISEN 625• Engineering Data Analysis - ISEN 613• Cognitive Systems Engineering - ISEN 631• Survey of Management - MGMT 655• Financial Management - FINC 635

6

COURSES:

Texas A&M University – Course Projects

• FABMS (Flexible Axle Beam Manufacturing System) - ISEN 641

– Introduction to the AS – IS system

– Functional and Operational models Description

– The Business Case – Problem Description

– Solution concepts definitions & Rationale

– CONOPS

– Dispatch Plan and sequence Algorithm

– Simulation studies

7

FABMS (Flexible Axle Beam Manufacturing System) - ISEN 641

• FABMS (Flexible Axle Beam Manufacturing System) - ISEN 641

– Performed root cause & business case analysis for the identified feasible solutions to infer the optimum solution.

– Simulated the in feed queuing system using SIMIO and VENSIM to understand system dynamics.

– Improved capability of the manufacturing system and increased the productivity of the system.

8

Shared Mental Models in Sport Teams – ISEN 689

• Objective

There has been a rise in the use of analytics in sports over the past few years. Much of the existing research in this area has focused on:

– Individual and team performance. Analysis of successful teams to formulate a winning theory.

– Various studies trying to come up with new or better ways to measure performance (Basketball Analytics).

– Research on shared mental models - has been applied in various contexts ranging from tennis.

9

Source - http://cogaids.stanford.edu/whatarecogaids.html

Shared Mental Models

Basketball Analytics

Individual & Team

Performance

• Focus of the research

• Literature Review• Shared Mental Models• Individual and Team Performance• Basketball Analytics

Shared Mental Models in Sport Teams – ISEN 689

• Measuring Shared Mental Models

The data collection methods that will be used for this research work :

• Observations – Each game will be videotaped and later observed thoroughly to track player movement.

• Surveys – Players will be asked to complete a post and pre session survey.

• Interviews – Players will be interviewed after games to understand the surveys better.

This research builds on the hypothetical model used in “Shared Expectations

and Implicit Coordination in Tennis Doubles Teams”. Some of the measurable include:

• Team Familiarity

• Task Experience

• Coaching Received

These measurable lead to:

• Shared Knowledge

• Implicit Coordination10

Shared Knowledge

Coaching Received

Team Familiarity

Task Experience

Depicted above is of questions from the survey

Shared Mental Models in Sport Teams – ISEN 689

• Anticipated Outcomes

– This research will give us more insight into shared mental models, and the impact it has on a team’s performance.

– This research will present a way to measure a team’s Shared Knowledge, Implicit Coordination and Socio-Technical Congruence.

– This research will show the effect of these measurable on the performance of a team.

– The goal of this work is to contribute to the literature on shared mental models and team performance. This research will build on existing work on sports teams to understand the cognitive factors that influence team members’ ability to work together effectively as a team.

• Future Work

– This research can be applied in pro basketball leagues and can used to improve team cognition and strengthen team work.

– This research will serve as an useful tool for coaches and teams understand their teammates better, identify gaps in understanding.

– This research will contribute as basis for further research on other pro sport leagues.

– This research might also help identify other factors that contribute to team cognition. These factors could further lead us to understanding shared mental models better.

– It would also help real-world sports teams to better understand the cognitive dimension teamwork and performance.

Recognizing Bench Talent in the NBA using R -ISEN 613

• Objective and Motivation

– The key objective of the project is to identify players who can come off the bench and perform efficiently for a team.

– This project aims to provide information on identifying player patterns to aid coach’s decision on substitutions and ensuring that the right player is brought on court.

– In addition, this also helps in identifying new and undiscovered talent.

12

• Proposal Statement

– Build a model to group the current NBA player’s accordance to their abilities and performance. The model classifies bench players with similar skillset to the starters.

Recognizing Bench Talent in the NBA using R -ISEN 613

13

• Approach

– Data Imputation

– Identify important parameters

– Develop model

– Results and Interpretation

• Data SetThe NBA data set that was employed in this project was pulled from NBA.com. The various tables used are described below:

- Defense: how well the players guard another player.Data Set: Opponent Shooting

- Offense: when guarded by a defender. Data Set: Shooting Data

- Clutch/Shot Clock: Player’s ability to deal with time pressure. Data set: Shot Clock Data

Recognizing Bench Talent in the NBA using R -ISEN 613

• Data Imputation

– Since the parameters in the dataset are large, we use PCA (Principal Component Analysis) to reduce the number of factors which influence clustering.

– Plotting the graph for the proportion of variance versus the principle component clearly shows a sharp decline in the variation after the third principle component, hence only the first three principle components are taken to represent the data accurately.

– The PCA also works to eliminate the co-relation with various data sets.

14

• Identify Important Parameters - By doing PCA analysis, the number of important parameters are identified when we see a sharp decline in the variation. This also eliminates the co-relation with various data sets.

Recognizing Bench Talent in the NBA using R -ISEN 613

• Perform K-Means

- Depending on the number of important principal components, the players can be clustered into various clusters.

- The number of clusters is determined through trial and error.

15

Recognizing Bench Talent in the NBA using R -ISEN 613

16

Player Name Min/Game Position Cluster Number

1. Eric Moreland 0.7 PF 2

2. Russell Westbrook 34.4 PG 4

3. Kevin Durant 33.8 SF 8

4. James Harden 36.8 SG 4

5. Stephen Curry 32.7 PG 2

6. DeMarcus Cousins 34.1 C 2

7. LeBron James 36.1 SF 8

8. Klay Thompson 31.9 SG 4

9. Carmelo Anthony 35.7 SF 2

10. Dwyane Wade 31.8 SG 3

Offensive Player Efficiency Rating

Results and Interpretation• We use the offensive

efficiency rating from NBA.com

• Respective cluster for the players are identified.

Recognizing Bench Talent in the NBA using R -ISEN 613

17Results from the Clustering Analysis

Player Name Min/Game Position Cluster Number

1. Reggie Williams 5.3 SF 8

2. Shayne Whittington 5.4 PF 2

3. Brendan Haywood 5.4 C 2

4. Russ Smith 5.4 PG 4

5. Jack Cooley 5.4 SF 2

6. Andre Dawkins 5.5 SG 4

7. Nazr Mohammed 5.6 C 8

8. Steve Novak 5.6 SF 4

9. Kalin Lucas 6.0 PG 2

10. Jerel McNeal 6.0 SG 2

Results and Interpretation• The players who belong to

the same cluster can be used as a substitute for resting the starters who are in the same cluster.

• We consider players who have played more than 5 minutes a game to improve the accuracy of the model.

• This is done for the other data sets as well.

Point-of-Sale System Analysis – ISEN 631

• Scope

– Interactions between server-system

– Environmental factors (lighting, noise, etc.)

• Goals

– Improve interactions between server-system

• Measured by lower interaction time & fewer errors

• Technology

– Touch screen interface.

18

Point-of-Sale System Analysis – ISEN 631

• Task AnalysisFour key tasks for providing customer service:

1. Take the Order – Note down customers’ orders.

2. Fulfill Order – Bring the ordered food items to the customer.

3. Check for further orders – Receive customers’ other orders if any.

4. Print and Process Check – Print the check when the customer is ready and process the payment.

• Data Collection Methods– Unstructured Interviews

– Think-aloud verbal protocol

– Contextual Inquiry

• Took on ‘apprentice’ role

– Direct Observation

• Focused on three main areas– Inputting order

• Avg time = 10 sec

– Printing check

• Avg time = 13 sec

– Processing payment

• Avg time= 7 sec(cash), 19 sec(debit), 11 sec(credit)

19

Point-of-Sale System Analysis – ISEN 631

20

Order

Server

1 2 3 4 5 average mistake

A 11s 15s* 10s 9s 13s 11.6s 1*

B 10s 8s 12s 11s 10s 10.2s 0

C 8s 9s 11s 9s 11s 9.6s 0

D 9s 10s 10s 11s 10s 10s 0

Order

Server

1 2 3 4 5 average mistake

A 13s 15 12s 12s 13s 13s 0

B 12s 13s 12s 14s 10s 12.2s 0

C 11s 10s 11s 15s 11s 11.6s 0

D 11s 14s 15s 11s 13s 12.8s 0

Table 1: Time to input order for different servers with average time and mistake tally

Table 2: Time to print check for different servers with average time and mistake tally

Point-of-Sale System Analysis – ISEN 631

Problems and Solutions1. Splitting Checks - The system does not allows the servers to split the check by the customers’ individual orders.

• Split check must be tabulated manually

• Leads to mistakes

• Time consuming, especially when the restaurant is loud and/or crowded

– Implement system design to easily allow this function.

21

System implemented at IHop

2. Servers do not know when food is ready in the kitchenOrders are linked to server ID

• System lets the servers know when the food is ready.Provide servers with pager system for kitchen to announce when food is ready

Point-of-Sale System Analysis – ISEN 631

Problems and Solutions

3. Include pictures of food items – Better representation to help servers.

22

4. Low visibility of which food item is selectedRedesign with neutral background color Color changes when item selected

Possibly give haptic feedbackImportant during high background noise times (i.e. karaoke night, busy weekend evenings)

Point-of-Sale System Analysis – ISEN 631

Problems and Solutions

5. Servers have to recall pin number each time they access system.

• Smart card/swipe card login

• Use pin as backup method

23

6. Reach for the sky solution• Integrate menu/payment/pager system into

tablet based computer.• Facial recognition.• Eliminates need to write order down.• Order completion notification.• Ability to process payment at the table.

Thank You!

24