Skip to content

To increase transparency and efficiency in intralogistics, we are looking for a solution to classify moving boxes in a production area. By using the latest 2D-Lidar technologies powered by Pepperl+Fuchs, you will generate valuable insights. Get creative and pull out anything of your machine learning toolbox!

Notifications You must be signed in to change notification settings

expectopatronm/Pepperl-Fuchs-Lidar-Classification-Challenge

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 

Repository files navigation

Pepperl+Fuchs as well as certainly many other hardware manufactures are implementing new Industry4.0 and IIoT concepts in their production to make their business overall more profitable. Whereas in the actual manufacturing process many - if not all - steps are already automated, the full potential for further automation in intralogistics is still to be leveraged.

At Hackdays Rhein-Neckar 2021, we are looking for a smart solution in order to monitor the flow of materials within such a manufacturing process. Automatic transportation of boxes and carriers are often done by conveyor belts which guide the goods to their destination. To increase transparency and efficiency in this process, we are looking for a Lidar-driven solution to autonomously monitor these belts..

The Challenge

At any given time, the monitoring system shall give an overview of how many and what kind of boxes have been transported on a certain belt. By using 2D Lidar Scanners by Pepperl+Fuchs we provide you access to the latest laser technologies when it comes to distance measurements. The R2000 is a 2D Lidar scanner with a 360°-angle of measurement making use of pulse ranging technology and is capable of angular divergence of 10 mrad transversal and 2mrad longitudinal. Find the most suitable way to classify the boxes that have been on a certain belt and get creative


The Experimental Setup

Visualizing Lidar Data

3D Software Render

Surface Plot

Initial K Means clustering of single scan data to determine seperability of features

Method 1: RBF Resampling + Support Vector Machines

Classification Accuracy: 43%

Method 2: PoinNet Point Cloud Classification

Live Stream Data

About

To increase transparency and efficiency in intralogistics, we are looking for a solution to classify moving boxes in a production area. By using the latest 2D-Lidar technologies powered by Pepperl+Fuchs, you will generate valuable insights. Get creative and pull out anything of your machine learning toolbox!

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published