“I have worked under Abhinav's leadership for 18 months in SLB. During this time Abhinav was always pursuing cutting edge technologies to resolve challenging business problems, with consistent results in the form of published papers and business savings. His style of management is smooth and assertive, always guiding the team for excellence and professional growth.”
About
My org is building Lumi™…
Contributions
Activity
-
Would you rather spend ₹50 lakh on your education abroad and earn ₹1cr abroad after graduation OR study for free and come back to India on Indian…
Would you rather spend ₹50 lakh on your education abroad and earn ₹1cr abroad after graduation OR study for free and come back to India on Indian…
Liked by Abhinav Kohar
-
Isn’t it wild how all top universities like Stanford, MIT, IIT etc. offer FREE online courses for real learners, but also have pricey executive…
Isn’t it wild how all top universities like Stanford, MIT, IIT etc. offer FREE online courses for real learners, but also have pricey executive…
Liked by Abhinav Kohar
-
💙 Welcome to McKinsey Los Angeles, located in the heart of downtown. It's the second-tallest building west of the Mississippi River. Our team here…
💙 Welcome to McKinsey Los Angeles, located in the heart of downtown. It's the second-tallest building west of the Mississippi River. Our team here…
Liked by Abhinav Kohar
Experience
Education
-
University of Illinois at Urbana-Champaign
MS CS (Thesis)
Thesis : Artificial Intelligence for high frequency, non-uniform, high dimensional time series data -
Licenses & Certifications
Volunteer Experience
Publications
-
Condition Based Maintenance of Oil Field Cement Pumper: A Data Driven Approach
SPE Western Regional Meeting, 2024, SPE-218870-MS
In this paper we describe the method and application of the four PHM models we developed and implemented in 2023. The first three distinct models monitor the performance of three different centrifugal pumps designated for distinct functionalities within our system. The fourth model monitors the performance of the radiator that dissipates heat from the coolant that regulates the temperature of lubrication systems for the engine and transmission. All four models were tested on historic data and…
In this paper we describe the method and application of the four PHM models we developed and implemented in 2023. The first three distinct models monitor the performance of three different centrifugal pumps designated for distinct functionalities within our system. The fourth model monitors the performance of the radiator that dissipates heat from the coolant that regulates the temperature of lubrication systems for the engine and transmission. All four models were tested on historic data and successfully deployed to identify deviation from healthy operating zones in the production jobs. The results are promising, given the models have identified 22 defects since deployment.
The Dataiku platform was used for data processing, analysis, and algorithm development for models. The first three models were developed using a polynomial regression method along the root mean square error (RMSE) metric. The fourth model was developed using a curated dataset to delineate the zone of interest and to define the thresholds for detecting deviations. The results of the PHM models were visualized on interactive dashboards, statistically significant outliers are analyzed in real time, and used to alert the operations and maintenance teams in the field.Other authorsSee publication -
MWD Tools’ Electronic Components Data Driven Failure Detection,
2024 SPE Western Regional Meeting, SPE-218834-MS
Within the oil and gas industry, measurement-while-drilling (MWD) tools have several sensors to provide telemetry data that allow performing either live or post-job health diagnostics of the tool components and effectively replace/repair them to avoid unnecessary nonproductive time. Nevertheless, health checks are often purely telemetry-based, not using past job information to quantify cumulative component degradation, which increases the probability of failure during subsequent…
Within the oil and gas industry, measurement-while-drilling (MWD) tools have several sensors to provide telemetry data that allow performing either live or post-job health diagnostics of the tool components and effectively replace/repair them to avoid unnecessary nonproductive time. Nevertheless, health checks are often purely telemetry-based, not using past job information to quantify cumulative component degradation, which increases the probability of failure during subsequent jobs.
This work proposes a methodology that (1) introduces historical data to improve on failure detection and (2) benchmarks it against single-job-based models over the same features for a particular MWD tool. We find that by stacking historical features (as multivariate time series), failure prediction using one-dimensional convolutional neural networks provides better estimates of the probability of failure.
We apply our methodology to five of the most common electronic component failure detection models trained over the time series the telemetry channels of 38 coiled tubing drilling MWD assets, totaling 752 jobs, and present the comparison between single-job-based models (with an average F1-score 0.66) and historical data-based models (with an average F1-score of 0.75).Other authorsSee publication -
Mitigating Nonproductive Time: A Novel Algorithm for Dsl Fault Detection
IPTC
Digital slickline (DSL) has been introduced to improve the efficiency of intervention operations in both onshore and offshore wells. DSL cables provide a real-time two-way-telemetry path between the acquisition system at the surface and the downhole sensors. Most nonproductive time (NPT) in DSL operations stems from telemetry issues due to cable faults despite the system's robustness. To address this, we developed a data-driven framework for identifying potential cable damage and its…
Digital slickline (DSL) has been introduced to improve the efficiency of intervention operations in both onshore and offshore wells. DSL cables provide a real-time two-way-telemetry path between the acquisition system at the surface and the downhole sensors. Most nonproductive time (NPT) in DSL operations stems from telemetry issues due to cable faults despite the system's robustness. To address this, we developed a data-driven framework for identifying potential cable damage and its approximate location using the DSL logging telemetry data, including communication signals, pressure, and depth. We tested our method on 992 real-case downhole jobs across almost 30 countries. To validate our method, we compared the method predictions for 60 jobs with labeled potential faults (i.e., cable damage), reaching an accuracy of 98% when considering whether the job has a fault. Thus, our framework enhances cable management, reducing NPT and associated costs.
Other authorsSee publication -
Deep learning models for high-frequency financial data
IDEALS
The limit order book of a financial instrument represents its supply and demand at each
point in time. The limit order book data can be used to predict the future price of the
financial instrument. We develop deep learning models to capture the high dimensional
data distributions (on Rd) of the limit order data. These models exploit the underlying
structure of this complex data. We develop a uniform data grid model for limit order book
data to achieve state-of-the-art accuracy…The limit order book of a financial instrument represents its supply and demand at each
point in time. The limit order book data can be used to predict the future price of the
financial instrument. We develop deep learning models to capture the high dimensional
data distributions (on Rd) of the limit order data. These models exploit the underlying
structure of this complex data. We develop a uniform data grid model for limit order book
data to achieve state-of-the-art accuracy for predicting price changes in a stock. We also
develop a novel way to use non-uniform events from the limit order book data to train a
non-uniform grid data model. This model substantially and consistently outperforms our
uniform data grid model. Both the models have been trained and tested over a wide range of
periods spanning multiple years for many stocks. The out-of-sample predictions are stable
across time for both the models as shown by tests for multiple stocks. Given the huge size
of the dataset we use a cluster of CPUs and GPUs to perform our experiments.
iiOther authorsSee publication -
Support for power efficient mobile video playback on simultaneous hybrid display.
IEEE Computer Society
Mobile devices, such as smartphones, e-books, and tablets, have limited battery capability because of the constraint of battery size and mobility requirement. However the large color displays on those devices put more tensions on this situation as the displays consume a large portion of the total battery power. A TOLED-EPD hybrid display that integrates a transparent OLED (TOLED) with an electrophoretic display (EPD) has been emerging to reduce the energy usage of displays. The technology…
Mobile devices, such as smartphones, e-books, and tablets, have limited battery capability because of the constraint of battery size and mobility requirement. However the large color displays on those devices put more tensions on this situation as the displays consume a large portion of the total battery power. A TOLED-EPD hybrid display that integrates a transparent OLED (TOLED) with an electrophoretic display (EPD) has been emerging to reduce the energy usage of displays. The technology displays information selectively on one of the displays based on the update rate of content, thus reduces the energy usage. In this paper, we propose a design of mobile video playback, Decoder4Hybrid, for the hybrid displays. The proposed approach supports encoded video playback based on the update frequency of each block, which is exploited by the hybrid display controller to determine which display should be used to show a MPEG encoded block. A fast DCT-based heuristic algorithm is proposed to detect the changes between frames at block level with minimal computation cost. Experimental results show that the proposed approach can save up to 40% power with acceptable video quality.
Other authorsSee publication
Projects
-
Byzantine fault tolerance and large scale graph processing on hadoop
-
Honors & Awards
-
Siebel Scholar
Tom and Stacy Siebel Foundation
https://en.wikipedia.org/wiki/Siebel_Scholars
The Siebel Scholars program was established by the Thomas and Stacey Siebel Foundation in 2000 to recognize the most talented students at the world’s leading graduate schools of business, computer science, bioengineering, and energy science. Each year, more than 90 graduate students at the top of their class are selected during their final year of studies based on outstanding academic performance and leadership. -
President of India Gold Medal
Shri Pranab Mukherjee, President of India
-
IIT Patna Institute Gold Medal
Ajai Chowdhary, Founder HCL, Chairman IIT Patna
-
State Bank of India Award
State Bank of India
-
DAAD-WISE Scholar
German Academic Exchange Service
Deutscher Akademischer Austauschdienst or German Exchange Service is one of the most coveted internship which provides 2-3 months internship opportunity at universities in Germany. 100 scholars selected all across India.
The selection is based on academic achievements, leadership potential, and candidate's overall profile.
https://www.daad.de/deutschland/stipendium/datenbank/en/21148-scholarship-database/?daad=1&detail=50015295&origin=4&page=1&q=wise&status=1&subjectGrps= -
Excellence award
Kapil Sibbal, Minister Human Resource & Development, Government of India
-
Silver Medal, Indian National Mathematics Olympiad
-
Test Scores
-
GRE
Score: 329/340
-
TOEFL
Score: 116/120
-
IIT JEE Mains
Score: All India Rank 729
Out of 1.1 Million candidates (General Category)
Recommendations received
10 people have recommended Abhinav
Join now to viewMore activity by Abhinav
-
✅ I will be speaking at 2nd Generative AI week (Atlanta) 𝐏𝐫𝐞𝐬𝐞𝐧𝐭𝐚𝐭𝐢𝐨𝐧: 𝐎𝐩𝐞𝐫𝐚𝐭𝐢𝐨𝐧𝐚𝐥𝐢𝐳𝐢𝐧𝐠 𝐚 𝐍𝐞𝐱𝐭-𝐆𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐨𝐧…
✅ I will be speaking at 2nd Generative AI week (Atlanta) 𝐏𝐫𝐞𝐬𝐞𝐧𝐭𝐚𝐭𝐢𝐨𝐧: 𝐎𝐩𝐞𝐫𝐚𝐭𝐢𝐨𝐧𝐚𝐥𝐢𝐳𝐢𝐧𝐠 𝐚 𝐍𝐞𝐱𝐭-𝐆𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐨𝐧…
Liked by Abhinav Kohar
-
Two of my friends at openAI looking to leave. Very interesting dynamics at play. Work is great, teams are high performing, but saw this at Google[x]…
Two of my friends at openAI looking to leave. Very interesting dynamics at play. Work is great, teams are high performing, but saw this at Google[x]…
Liked by Abhinav Kohar
-
The MSR Undergrad Intern Program is now accepting applications for 2025. The program goals are to provide undergraduate students with research…
The MSR Undergrad Intern Program is now accepting applications for 2025. The program goals are to provide undergraduate students with research…
Liked by Abhinav Kohar
-
Exciting News! #AWS Energy is thrilled to announce a new collaboration with SLB and Vista as part of a Memorandum of Understanding (MOU) to…
Exciting News! #AWS Energy is thrilled to announce a new collaboration with SLB and Vista as part of a Memorandum of Understanding (MOU) to…
Liked by Abhinav Kohar
-
We are seeking an experienced Staff Software Engineer based in the United States, with a minimum of 10 years of software development experience…
We are seeking an experienced Staff Software Engineer based in the United States, with a minimum of 10 years of software development experience…
Liked by Abhinav Kohar
Other similar profiles
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore More