Artifical Pancreas for Type-1 Diabetes Mellitus Patients
- Working under Dr. Radhakant Padhi to implement control algorithms in Aritifical Pancreas.
- Type-1 DM patients suffer from hyperglycemia and hypoglycemia.
- This may also result in diabetic neuropathy, diabetic retinopathy and diabetic nephropathy.
- Aritifical Pancreas can help regulate the glucose level in the blood by injecting insulin subcutaneously into the blood stream.
- In order to control the amount of insulin to be injected, a model predictive control algorithm is being developed.
- Model Predictive Control requires an accurate model that captures glucose-insulin dynamics in the patient.
- Completed parameter estimation for Gluco-regulatory model, Subcutaneous Insulin Kinetics model and Meal model
- Another requirement to implement model predictive control is state information. In real time implementation, only Glucuose measurements are available.
- Using these glucose measurements, which also happen to be noisy and delayed, other states of the system are to be estimated using an Extended Kalman Filter.
Application of Control Theory in Cyber Physical Systems
- I completed an independent research study under Dr.Justin Ruths at UT Dallas.
- Developed a code to simulate a vehcile platoon. This was achieved through adaptive control.
- Applied control algorithms in a network of RC cars. I implemented PD control in python on Anki OverDrive cars using Anki's SDK
which is compiled in a Linux Environment. The cars were made autonomous by running the SDK code and the python code on a Raspberry Pi and achieved vechile platooning using an Ultrasonic Sensor.
Tata Advanced Systems Ltd., Senior Executive (Flight Control Systems)
Project: Autopilot Development for Medium Altitude Long Endurance (MALE) aircraft
- Contributed in the finalization of System Specification Document for Automatic Flight Control System (AFCS).
- Trimmed non-linear 6 D.O.F aircraft model at various envelope points and linearized the model at those points to obtain linear state space model and their transfer functions.
- Obtained open loop longitudinal and lateral characteristics of the aircraft at various points in the envelope.
- Developed control laws for Stability Augmentation System (SAS) in order to improve the short period, dutch roll, spiral and roll characteristics of the aircraft to meet MIL - 8785 C requirements.
- Designed attitude hold loops (pitch and roll hold modes) as well as coupled outer loops (airspeed hold, altitude hold and heading hold modes) to meet settling time requirements as per MIL - 9490E.
- Gain values for inner loops and outer loops were designed to meet Gain margin and Phase margin requirements as per MIL - 9490E.
- Completed Functional Requirement Structural Definition Document for Control Law Development (CLAW) for current version of aircraft model.
- Tested the control laws in Model-In-Loop (MIL) and Hardware-In-Loop (HIL) environment.
Project: Gimbal Stabilization of 10 kg Fixed Wing Drone
- Stabilized gimbal for Geo-pointing where camera points towards a specified location
- Improved target tracking by introducing a Kalman Filter to smoothen sensor values for camera stabilization.
- Implemented Fly-By-Gimbal functionality wherein the drone will follow the target specified by target tracking function.
Project: Swarm robots for disaster relief
- Devised a formation algorithm for a group of 10 follower drones to follow a team lead.
- Implemented an optimization algorithm wherein the drones in formation disperse and drop payload at various target locations and then assemble back in formation.
- Assisted in system design architecture for implementation.
ATS ELGI Ltd., Robot Programmer
- Programmed a FANUC 120iC 10L Robot that is incorporated in an Arc welding cell.
- Calibrated the Tool frame and created individual user frames for different positions for three different work stations.
- Implemented Position Registers and created the program in such a way that it is easily modifiable.
- Reduced the cycle time of the process by 18% from 76 minutes to 63 minutes.
- Improved the weld quality by increasing the fillet size, eliminated porosity, reduced undercut and spatter.
Difacto Robotics and Automation Pvt. Ltd., Trainee Engineer
-Performed material handling and spot welding operations using FANUC LR Mate 100iB and FANUC R2000iA 165F Robots respectively.
-Created off-line programs for material handling using ROBOGUIDE and automotive spot-welding using CimStation Robotics (CSR).
-Designed robot pedestals using the DriveWorks module in SolidWorks and created part models with sketches for fixtures using SolidWorks and AutoCAD respectively.
-Generated design concepts for a Robotic Ultrasonic Cutting and Crank Shaft Assembly Automation.
-Assembled fixtures and jigs that were utilized in Robot Spot Welding and Arc welding cells.