Design and Fabrication of a Magnetic Manipulator.

Parameterized the 3D space to select important geographical constraints.

Modeled the coil manufacturing Process and specified the optimization parameters.

FEM Modeling of the coils.

Selecting Electrical Equipment and developing Electrical circuits required for manipulation.

Designed Mechanical frames and boxes in Solidworks and fabricated using Lathe, milling, CNC, bending, and saw machines.

Data collection for AI algorithm development.

Created regression Models(ANN, ANN/SA, GEP) to the approximate magnetic field required for the actuation matrix.

Developed actuation logic.

This project was focused on stabilizing a Segway whit and whit out restrictions on inputs.

Deriving the model of Segway using inverse pendulum motion model and linearizing it around the desired position,

Implementing MPC controller to stabilize the segway with and without restrictions on input.

A 3D design of a Clutch System using Part, Surface, Assembly, sheet metal, and rendering.

Providing the analytical model of Sheet metal cuter.

Analyzing the displacement and rotation of the arm and cuter.

Analyzing the velocity of motion for a given torque.

PLC

- Supply tanks’ filling 

- Mixing 

- bottling 

The mixture is produced by mixing specific amounts from three different substances stored in the supply tanks. Afterward, the mixture is bottled and packaged for distribution.

Q- learning and Adaptive dynamic programming (ADP) to navigate a robot in the given environment starting from (2,3).

Development of Traceless Genetic Programming algorithm to predict the Shear Strength. This project included feature engineering, and the results were verified using modern statistical techniques.

Developing a single structure of ANN to predict the results of the following functions

y1 = x1 exp(−ρ)
y2 =cos(ρ)/x2
ρ = x1^2 + x2^2
Domain :
x1, x2 ∈ [−2, 2)

The project was challenging since the second function provides zero response and is less correlated to inputs than the y1.

Cleaning a data sheet with 78 inputs applying feature engineering techniques.

Introducing new features to enhance the performance of the algorithm.

Developing Ridge, Lasso, ElasticNet models to predict the House Price.