Responsibilities include performing research activities in the areas robotics, self-driving cars, AI, machine learning, and advanced control systems. Duties include the design and development of software engineering algorithms applied for collaborative robotics (Cobots) that are used in manufacturing, healthcare, and academic environments. Design artificial intelligent control algorithms for Advanced Driver Assistance Systems (ADAS) kits and mobile robotic platforms.

FIRSTSENSE MEDICAL (FSM) is developing a revolutionary breast cancer screening device with 96% projected sensitivity (detection rate) when combined with patient history and ultrasound in select cases. It is painless, radiation free and takes about 6.5 minutes and is done in a doctor's office.FSM has two models for screening. FIRSTSENSE Breast Exam® (FSBE) and SENTINEL. FSBE uses Advanced Thermal Profiling, Infrared and 3D cameras to view breasts from front and sides. The SENTINEL tester has a stationary IR camera and mirrors to view sides of the breast. Sophisticated proprietary software assists the clinician with analytical support. Both systems are qualified as a Class I device, with FDA 510(k) clearance as an adjunct to a mammogram. They use thermal images and a cold challenge analyzed by an in-house lab (1Q17 launch). FSM is also pursuing many clinical trials at major hospitals and universities -- St. Joseph Mercy at Pontiac and Ann Arbor MI, Univ. of Toledo, OH, and UCLA. My responsibilities include:The management of the software development activities of the bio-robotics system related to the medical image diagnosis, robot controls, sensor integration, embedded systems, cloud based data analysis, bio-statistics and user interactions. Directing the Software department staff, department budget, project proposals, requirements and go-to-market strategies. Involved in the product development of breast cancer tester, system architecture, functional aspects, specifications, use cases, V&V, HIPAA, FMEA, FDA level standards and documentation. Leading the FDA 510K clearance process and maintaining the IEC 62304 software standards and quality. I have filed over 3 patents and several trade secrets in areas of IR image filtering, autonomic nervous system response, CAD, etc.“Apparatus and method for detecting a presence of cancer,” Paul G. Angott, Pavan Vempaty, et al. US Patent 20150257650 and 20150257651 [Sept, 2015].

Takata is combining the active safety systems with its existing safety systems, such as seat belts and airbags, which it has developed over many years. Going forward, as the driving environment becomes more complex, the target is to think of all potential events that may occur around automobiles and move ahead with the development of Total Safety Systems.My responsibilities include:Design and development of monocular vision based Forward Collision Warning (FCW). Subsequently as a lead of the monocular vision group, I led the design and development of machine learning and detection methods based on Pattern Recognition, Wavelets, Support Vector Machines (SVM), and AdaBoost techniques. With the state of art embedded technologies and rapid prototyping tools, I assisted in deploying these applications as active safety products known as SafeTrakTM for North American and European truck corporations. I designed the algorithm based upon the ISO26262 quality standards for the driver assistance system that utilizes the vehicle’s GPS unit to assist the driver for effective on-road navigation. The algorithms utilize a database that contain information regarding the road and junction characteristics. Based on the GPS location, the driver is informed of the upcoming road curves, speed limits, roadside traffic signs, signals etc. Thereby increasing the on-road safety and driver alertness.I have filed over 9 disclosers and over 4 trade secrets in ADAS area, which include Autonomous Vehicle Navigation, Smart Vehicle Systems through Safety Digital Maps, 3D Augmentation of GPS turn by turn directions, Live Traffic Analysis Updates, Pedestrian and Vehicle detection, 2011-2013“Driver Assistant System,” Troy Cooprider, Mike Schmidlin, Faroog Ibrahim, Pavan K Vempaty et al. US Patent 20120303222, http://www.google.com/patents/US20120303222.

I worked as the systems design engineer at General Electric (GE) Transportation. During this time, I was involved in the algorithm development of GE locomotive’s smart fuel saving product called Trip Optimizer. Trip Optimizer provides the train’s pilot with the information on efficiently controlling the trains throttle, notch, speed based on the train’s location. By following the indications provided by the Trip Optimizer, the fuel efficiency is increased dramatically. For the Trip Optimizer algorithm, determining the precise location of the train is an important factor. A differential GPS is used for providing the algorithm not only the location of the system but also an estimated speed of the train. For this algorithm, I was involved in the design and development of an estimation control system based on Unscented Kalman filtering for precisely localizing the train’s position and speed estimation. The speed of the train is determined by fusing the GPS location and the train’s wheel speed (derived based on the wheel diameter). Further, I have used the Unscented Kalman Estimation filtering bias to identify and warn the train operator if wrong wheel diameter is entered Also, using the Model Predictive Control State Estimation methods, I have developed techniques to estimate location of the train when there is a partial absence of the GPS signal (ex: tunnels). “Estimation of locomotive wheel diameter through Kalam Filter Bias value,” Vishram Nandedkar, Pavan K Vempaty, Ajith Kumar, et al. US Patent 20130268172, http://www.google.com/patents/US20130268172.