Our Projects

Research

Our Projects (on going)

Swarm Robotics: Towards Decentralized Intelligence in Multi-Robot Systems

As shop floor activities shift from mass customization to mass individualization, Multi-Robot Systems (MRS) offer a flexible, adaptable, and cost-effective solution. These systems, whether Single Robot & Single Task (SRST) or Multi-Robot & Single Task (MR-ST), require efficient planning, resource allocation, task decomposition, and scheduling to optimize performance. Achieving decentralized intelligence involves developing algorithms for heterogeneous MRS, along with robust monitoring and control strategies for communication, coordination, and navigation. This enables MRS to autonomously adapt and fulfill customer-specific needs efficiently.

Vision-Based Quality Control Assurance

Quality assurance in Additive Manufacturing (AM) relies on real-time monitoring and process optimization to ensure part accuracy and reduce defects. Machine vision systems and sensors detect irregularities during production, enabling immediate corrective actions. Integrated machine learning algorithms analyze defect patterns and recommend optimal parameters, improving quality and efficiency. This proactive approach reduces waste, shortens inspection times, and enhances consistency, making AM ideal for precision industries like aerospace, automotive, and healthcare.

Research

Our Projects (Digital Twin)

Digital Twin for Diagnostics of Ball Bearing

Vibration analysis is key for maintaining rotating machinery like centrifugal pumps and gearboxes. In Industry 4.0, Digital Twins (DT) use physics-based models to develop and update virtual replicas of machines. For example, a ball bearing and shaft assembly model is created using finite element and discrete element methods, with vibration sensors mounted on the bearings for condition monitoring. The DT is calibrated with real sensor data, and machine learning algorithms, like Multi-Category Support Vector Machines (MSVM), are trained using this data for predictive maintenance. This approach optimizes performance and reduces downtime in industrial machinery.

Research

Our Projects (Digital Twin)

Virtual Twin of Smart Assembly Line

The Virtual Twin of Smart Assembly Line leverages cutting-edge technologies to create a digital replica of a customized manufacturing facility. This innovative platform integrates Omron's COBOT TM5, Automated Storage and Retrieval System (ASRS), Omron's Viper, and Omron's SCARA robots with Chain Conveyor to perform diverse manufacturing tasks.

The Smart Assembly Line has been modeled utilizing Dassault Systèmes 3DEXPERIENCE CATIA, while 3DEXPERIENCE DELMIA enables realistic and dynamic simulation of manufacturing processes. This virtual twin enables:

  • Enhanced production planning and optimization,
  • Improved collaboration and training,
  • Increased efficiency and productivity, and
  • Real-time monitoring and analytics

By bridging the physical and digital worlds, this project paves the way for Industry 4.0 adoption, fostering smart manufacturing excellence and unlocking new possibilities for innovation and growth.

Research

Our Projects (Human-Robot Collaboration)

Robots Movement in Human Spaces: Dynamic Path Planning and Navigation

This project focuses on enabling robots to navigate safely and efficiently in human-centered environments like factories, hospitals, and homes. Traditional path-planning algorithms are inadequate for dynamic settings, so the project develops real-time, adaptive algorithms that use sensor data, machine learning, and mapping technologies to avoid collisions and maintain a comfortable distance from people. The challenge is balancing computational efficiency with safety and human comfort, avoiding sudden movements while ensuring efficiency. Integrating SLAM for environment mapping and reinforcement learning for adaptive decisionmaking, the project aims to create robots that seamlessly interact in shared spaces.

Object Affordance and Actions

The maintenance and service sectors are highly valued pillars of product lifecycle management. Servicing products or warranty claiming is essential for a company's growth. As of today, the disassembling of the products for viewing and maintenance is done manually with the help of a skilled workforce. However, many challenges (e.g., skilling of the workforce, heterogeneity of products) need to be addressed to maintain the required level of service. Thus, it is necessary to automate the removal of screws to reduce non-value-added tasks without human intervention. The research project aims to develop an automation technology for disassembling various screw patterns using image-based recognition and affordance-based physical manipulation of robots for industrial applications.

close
close
APPLY NOW APPLY NOW

Admission Help

Admission Help

+91-836-2378103

close
close
phone
mail

info@kletech.ac.in

coe@kletech.ac.in (Controller of Examinations)

Chat now Chat now