Portofolio

Autonomous collaboration during a demonstrated task

Flexible task completion with human collaboration:

We show how an end goal can be demonstrated to a Robot and then left to complete the task even when there a multiple start points in the task

Autonomous completion of a demonstrated task

Flexible task completion

We show how a robot can complete a task when given an end goal by a human to strive for. This opens the doors for flexible and rapid programming of robot for flexible packaging of goods. 

Digitizing manual assembly tasks

Vision based tracking of manual tasks

We demonstrate how a task can be tracked using Vision based AI Algorithms. The technique we are using does not require too rigid setups making it very adaptable to various settings. This technique can be used to create a digital thread and record of a manual process. 

Digitizing manual assembly tasks

Vision based tracking of manual tasks

We demonstrate how a task can be tracked using Vision based AI Algorithms. The technique we are using does not require too rigid setups making it very adaptable to various settings. This technique can be used to create a digital thread and record of a manual process.

Robot assisted Assembly of Parts

Human-collaboration on assembly tasks

In this video, we show our architecture that is capable of enabling a collaborative robot to infer a human's action and support the assembly task that a human is carrying out. 

Autonomous Landing on moving platform

Vision based landing on a moving platform

Here we show how an Unmanned Aerial Vehicle could autonomously land on a moving platform using only visual information to track the moving platform and make decisions. 

Digital Twin of a Train maintenance Depot

Digital Twin of a maintenance factory

Here we develop a medium fidelity immersive environment model of an actual factory. We made use of medium fidelity in order to ensure that real time computation is possible within time constraints and to ensure that users derive full benefits of the immersive environment without lags. We applied a combination of Laser point cloud, drawings and field studies to obtain the dimensions of the artefacts in the factory as well as the factory itself.

Digital Twin of a Train maintenance Depot

Digital Twin of a maintenance factory driven by Discrete Event Simulation 

Here we develop an immersive environment model of an actual factory. Using cutting edge computer logic, we connected  a Discrete Event Simulator to the immersive environment. This enabled us to be able to visualise material flows and human flow in a factory without actual deployment. This saves costs in terms of rework and also enables engineers to explore alternative processes that could be more efficient without actual deployment in the real world. 

AI Website Generator