Research Papers and Seminars

Building Knowledge through Research and Sharing through Seminars

As a part of knowledge sharing across the organization, we actively focus on the following:

International Conferences
Tech Journal - TechTalk@KPITCummins
Lecture series by eminent scientists

International Conferences:

• “Business Process Maturity Model – Tool to Assess Process Health" by Mahesh R Jadhav and Gauri Sapre in the 11th International Conference on Informatics and Semiotics in Organizations (ICISO 2009), 10-11 April, 2009, Beijing - China.
• “Novel Approach for ROI Extraction in Vision based Vehicle Detection System” By: Chandrashekhar Padole in IEEE International Conference on Digital Image Processing, ICDIP’09, March-2009, Bangkok , Thailand.
• “Security and Protection aspects of Automotive Architectures for external Influences” by: Nagabhushana B.S., Convergence 2008, Detroit, MI
• "Image Restoration using Wigner Distribution for Night Vision System”, by: Vinay Vaidya and Chandrashekhar Padole., 9th International Conference on Signal Processing (ICSP'08) Oct.26-29 2008 Beijing, China.
• “Framework to Aid Process Maturity Model Definition for Product Development in Global Environment” by: Somashekhar R.H., Convergence 2008, Detroit, MI.
• "Night Vision Enhancement Using Wigner Distribution" By : Vinay G. Vaidya and Chandrashekhar P., IEEE ISCCSP 08, Malta, Europe, March 2008.
• "Simulation Study of the battery system for automotive traction applications" By: Nagbhushan, B.S., SAE World Congress 2008, Detroit, Michigan. April 14-17, 2008.
• "Fuzzy Logic based Handoff Algorithm for Wireless Network" By: Chandrashekhar P., Presented in International conference on RF and Signal Processing Systems, RSPS-2008, India, 1-2 Feb, 2008.

Tech Journal - TechTalk@KPITCummins

TechTalk@KPITCummins is a technical Journal published by the CREST team every quarter. There are a variety of articles in this journal contributed by KPites and our esteemed customers. TechTalk@KPITCummins is not only a repertoire of knowledge, but the journal also serves as a platform wherein our customers and employees come together to explore and share their thoughts about the latest in technology. The journal has been well received by everyone and we hope that in the coming years it will serve as a benchmark for technical journals.

• Virtual Reality Vol.1 Issue 1, Jan - Mar, 2008
• Modeling & Simulation Vol.1 Issue 2, Apr - Jan, 2008
• Machine Interfaces Vol.1 Issue 3, July - Sept, 2008
• Data Security Vol.1 Issue 4, Oct - Dec, 2008
• Vision Based Automation Vol.2 Issue 1, Jan - March, 2009
• Concurrent Computing Systems I Vol.2 Issue 2, April - June, 2009
• Sensor Network - Vol.2 Issue 3, July - Sept, 2009
• Concurrent Computing Systems II Vol.2 Issue 4, Oct - Dec, 2009
• TechTalk - Jan 2010
• TechTalk - Apr 2010
• TechTalk - July 2010

to receive a copy of any of the above issues of TechTalk please send us a mail at TechTalk.

Lecture series by eminent scientists

• Dr. Padmanabhan, a renowned Astro-Physicist, delivered a lecture on the “Weight of the Universe” on November 15, 2007
• Dr. R.A. Mashelkar delivered a talk on “Innovation as a way of life- Personal Reflections” in KPIT Cummins Infosystems Limited, Hinjewadi, Pune on February 11, 2008
• Dr. Richard Silberstein from Brain Sciences Institute, Swinburne University, Melbourne delivered a lecture on “Creativity and Brain” on June 24, 2008.

"Business Process Maturity Model – Tool to Assess Process Health" by Mahesh R Jadhav and Gauri Sapre in the 11th International Conference on Informatics and Semiotics in Organizations (ICISO 2009), 10-11 April, 2009, Beijing - China.

Abstract: This paper introduces the Business Process Maturity Model (BPMM), a Model which we have designed and developed to assess the business process maturity. BPMM serves as a very good tool for assessing process maturity and readiness before investing huge amount into process management initiatives, certifications. BPMM assess Process Maturity, in other words health of process to deliver the desired results in the desired manner. BPMM focuses on capability of factors which determine the health of the process. BPMM allows organization to grade its processes from “Informal” stage to “Optimized” stage. This also enables organizations to ascertain areas of improvement. This is designed considering the requirement of process owners, to assess process maturity. It is easy to use, apply and easy to interpret. We have used this BPMM at one of our client. This helped in identifying bottlenecks in effective execution of the process and helped to develop improvement plan.

"Novel Approach for ROI Extraction in Vision based Vehicle Detection System" By: Chandrashekhar Padole in IEEE International Conference on Digital Image Processing, ICDIP’09, March-2009, Bangkok , Thailand

Abstract: Advanced Driver Assistance system (ADAS) is becoming a more popular feature in vehicle from safety point of view with the aim of reducing injury and accident severity. With vision system, it is possible to detect the vehicle with intelligent processing. We have presented here a novel method to extract the ROI of possible candidate for vehicle presence with low resolution method. These extracted ROIs can be further processed to detect the vehicle more precisely to make vehicle detection system less computational complex. Experimental results obtained by our algorithm show zero FRR (False Rejection Ratio) which is the main concern in ROI extraction for vehicle detection system.

"Image Restoration using Wigner Distribution for Night Vision System”, by: Vinay Vaidya and Chandrashekhar Padole., 9th International Conference on Signal Processing (ICSP'08) Oct.26-29 2008 Beijing, China.

Abstract: Night vision systems have become an important research area in recent years. Due to variations in weather conditions such as snow, fog, and rain, night images captured by camera may contain high level of noise. These conditions, in real life situations, may vary from no noise to extreme amount of noise corrupting images. Thus, ideal image restoration systems at night must consider various levels of noise and should have a technique to deal with wide range of noisy situations. In this paper, we have presented a new method that works well with different signal to noise ratios ranging from -1.58 dB to 20 dB. For moderate noise, Wigner distribution based algorithm gives good results, whereas for extreme amount of noise 2nd order Wigner distribution is used. The performance of our restoration technique is evaluated using MSE criteria. The results show that our method is capable of dealing with the wide range of Gaussian noise and gives consistent performance throughout.

“Framework to Aid Process Maturity Model Definition for Product Development in Global Environment” by: Somashekhar R.H., Convergence 2008, Detroit, MI

Abstract: The speed of innovations and pressure for reducing cost has forced automotive industry to move towards standardization. This standardization enhances the ease and success of Global Product Development (GPD) from a technology viewpoint. However, for overall success, an effective process model shall, additionally, support GPD. The process needs of the present day GPD programmes and more so the future ones, heavily, challenge the prevalent process maturity models. This paper presents a process framework that can be adapted by the industry to achieve industry-wide process maturity for GPD.

"Night Vision Enhancement Using Wigner Distribution" By : Vinay G. Vaidya and Chandrashekhar P., IEEE ISCCSP 08, Malta, Europe, March 2008.

Abstract: Night vision has been a subject of interest for various sectors such as border patrol, automotive industry, defense, etc. Accidents occurring at night could potentially prove to be fatal. Thus, in the recent times, there has been a significant interest especially from the automotive industry for conducting research in night vision. In this paper, we propose a new method for enhancing images taken under poor light conditions. This method is a modified version of the Wigner Distribution. Wigner distribution is one of the time–frequency representations of a signal. In this paper, we have shown how the modified Wigner distribution can be effectively used for supplementing the driver’s vision during night and during snow conditions. The new method is evaluated using the Mean Squared Error (MSE) criterion. Quantitative results and visual inspection show the robustness and effectiveness of this algorithm.

Abstract: With increasing levels of sophistication and addition of functional modules in automotives, the electric payload on the vehicle battery is increasing. This is especially true for traction applications employing 36V DC battery source. Interfacing of the existing alternators for battery charging, controlling of battery charging circuitry, issues related to circulating currents among individual batteries, temperature compensation, etc, invariably warrant for the design and implementation of a strategic Battery Management System (BMS). Considering the complexities associated with the batteries, an analytical model of the entire system is desirable. As a part of an identified project, an analytical model of BMS for traction application was built and the model was iteratively tuned to achieve successful validation by a series of experiments. The prototype hardware was built to implement the power circuitry for the traction motor and the BMS application code was generated using the RTW and fused on to a custom built BMS ECU. The BMS ECU was integrated with the power circuitry and an end to end validation was carried out as the proof of concept.

Abstract: Increased usability and different functionality of mobile devices in 3G/4G network, apart from speech communication, has created a need for merging one wireless network with another, although the geographical area for existence of one network varies with the other. When a mobile device, already in a particular GSM network, enters into an area of wireless LAN (WLAN), it should be able to take advantage of the higher bandwidth and cheaper cost of WLAN. Today, various wireless access technologies are available such as GSM, CDMA, WLAN, Bluetooth, WiMax, HomeRF etc. In the 4G network, all these and other upcoming wireless technologies would be merged to form one network. This means that the handoff algorithm needs to be so intelligent that the user would not be concerned with the type of network being used by the mobile device and the communication nonetheless continues with efficient bandwidth, less power, low packet loss, cheaper usage cost, etc. Conventionally, the handoff algorithms are based on measurement of signal strength. As every access technology has different range of signal strengths, signal based handoff algorithm fails to use the heterogeneous network efficiently. In this paper, we have used fuzzy theory to decide initiating the handoff process. We have compared the efficiency of fuzzy based handoff system to the conventional handoff algorithm.