Prototypes & research projects
PROTOTYPES
Integrating AI-Driven Robust Control Algorithm with 3D Hand Gesture Recognition to track an Underactuated Quadrotor Unmanned Aerial Vehicle (QUAV).
Features:
3D Hand gesture based operation
Battery Operated
Can perform aggressive maneuvers.
Promising Autonomous flight technology
Underactuated Air Cushion Vehicle (ACV) on Carbon Fibre Structure
Features:
Data logging using SD Card
Remotely Operated Using RC Control (6-Channels)
Battery Operated
Can lift up the payload upto 50 grams
Shares the parameters using IoT Broker
Underactuated Aircushion Vehicle on Carbon Fibre Structure
Unmanned Under-water Vehicle (UUV) Prototype
Features:
Data and Video logging using SD Card
Wired-Operated
Shares the parameters using IoT Broker
Identify some of the sea-creatures
Unmanned Under-water Vehicle (UUV)
Sewerbot Design with Rudder Design
Features:
IoT Enabled
Monitors temperature, altitude and environmental pressure
Identifies the issues in sewerage line based on computer vision techniques
Clears the blockage using special rudder design operated by BLDC motor
Hazardous Gas Detection using MQ sensors
Sewerbot Design with Rudder Design
Simultaneously Localization & Mapping (SLAM) based UGV Design
Features:
IoT Enabled
Monitors temperature, altitude and environmental pressure
Localization and Mapping feature
Unmanned Ground Vehicle with SLAM
IoT and Computer Vision enabled Robotic Manipulator Design
Features:
IoT Enabled
Pick and drop the package as per color detections
Different Color Detection
IoT and Computer Vision enabled Robotic Manipulator Design
IoT enabled Solar Tracking and Monitoring (Prototype)
Features:
IoT Enabled
Tracks sun and produces maximum power readings
Shares data on IoT Broker "Connect Things"
IoT enabled Solar Tracking and Monitoring
Computer Vision based Crop Monitoring
Features:
Wheat growth detection as per color and height
Temperature and humidity sensors
Water Sprinkling mechanism as per temperature
Computer Vision based Crop Monitoring
Load Management System Using LabVIEW
Features:
Input current and voltage measurements
Output current and voltage measurements
Load management and switching
LabVIEW based Graphical User Interface (GUI)
Load Management System Using LabVIEW
Research Projects
Vision-based State Estimation of an Unmanned Aerial Vehicle
Principal Investigator: Dr. Saiful Azrin B M Zulkifli
Main Research officer: Muhammad Irsyad Sahalan
Co-Researcher: Ghulam E Mustafa Abro
Grant Name: International Collaborative Research Fund RG2021-0873
Partners
Universitas Muhammadiyah Surkarta Indonesia and Universiti Teknologi PETRONAS, Malaysia
Abstract:
This project aims to deliver an indoor logistics delivery drone without the use of global navigation satellite systems (GNSS). This requires an alternative approach to achieve accurate state estimation. Therefore vision-based state estimation is proposed for indoor navigation. Camera carries the advantage of being low cost, low power and reliable provided armed with advanced algorithms. This vision approach targets to utilize fiducial markers with monocular camera. When a marker is captured by a camera, its position and orientation with respect to the camera’s coordinate frame is determined based on its homography transformation. The pose of the camera, and hence the drone, in world coordinate can then be inferred from known markers poses on the provided map. The advantage of using such markers is that it can be easily detected using standard pattern recognition algorithms and conventional classification methods. The development of fiducial marker localization can be applied to various scenarios such as non human contact robot in hospital, load carriage in warehouse, surveilance in complexes, and etc.
Development of Range Enhancement Device for Personal Electric Mobility
Principal Investigator: Dr. Saiful Azrin B M Zulkifli
Team Members
Mohamad Mirza Sadiq, Ghulam E Mustafa Abro, Ahmad Zaid Syakir Mohd Yazsid
Independent Project
Abstract:
One of the biggest obstacles to the adoption of electric vehicles is their limited operating range per full charge of the onboard energy storage. This issue is known as range anxiety - people are reluctant to depend on EVs due to myriad reasons such as running out of stored energy, lack of charging facilities, and the time it takes to charge. Much research has been carried out to address this concern; some propose combining conventional and electric engines for the best of both worlds, others propose strategically placing charging stations, and still others propose using algorithms to determine the best route for the vehicle. This research attempts to implement a simple current limiting mechanism for existing personal electric mobility (PEM) equipment such as electric scooters, which will track its onboard battery state-of-charge (SoC) and produce a new signal that controls the level of current provided to the motor. This can help to limit energy depletion from the battery and thus extend the scooter’s range. By using MATLAB-Simulink software, this study simulates a PEM and the current limiting mechanism. With the use of Arduino circuitry, this research involves tapping into the existing electric scooter to record and analyze serial communication between the user interface and motor controller, to limit the motor’s current and achieve longer range.
Hand Gesture-based Control of Autonomous UAV for Mobile Disinfectant Dispenser
Principal Investigator: Ghulam E Mustafa Abro
Grant Name: IEEE HAC & SIGHT Grant No: 22COVID20
Team Members
Khairel Danish Khairil Anwar, Dr Saiful Azrin B M Zulkifli, Dr. Vijanth Sagayan Asirvadam,
Abstract
Without a doubt, autonomous unmanned aerial vehicles (UAVs) will increase quickly and be used more and more in the coming decades. These unmanned aerial vehicles, also known as drones, are either designed to help people or perform tasks that involve people. These drones have grown to be faster, easier, and less expensive by integrating several technologies and supported by hybrid algorithms. These machines can also do a variety of tedious, filthy, challenging, and hazardous tasks. The deployment of machine learning- or artificial intelligence-based algorithms that can give drones today an autonomous flight mode and computer vision capabilities can be seen in the area of combining various technologies and hybrid algorithms. In order to stop the spread of COVID-19 among the crowd, this topic has been studied. It shares the prototyping and development of an autonomous quadrotor UAV based on machine learning and has a potential to maneuver and detect object. In order to identify the hand motion and attempt to keep a certain distance from the human, our proposed algorithm would stimulate the drone. Spraying disinfectants on the user's hands is the main idea. This research is one of the ways to make such autonomous UAVs for commercial usage in a time when everyone needs to take precautions to protect themselves against COVID-19.