Process robotics data with HPCC Systems

Owned by Lorraine Chapman
Last updated: Nov 11, 2021
2 min read

This project was completed during the 2020 HPCC Systems Intern Program by Jack Fields, American Heritage School of Boca/Delray, USA.

Find out about the HPCC Systems Summer Internship Program.

Resources Available to Learn More about this completed project:

Project Description

Process Robotics data with HPCC Systems Platform. Previous intern project by American Heritage High School (AH) student Aramis developed ROS package to collect data from various sensors and spray to HPCC Systems Thor. Aramis also created some ECL code to graph some data with HPCC Systems libraries. In this project we try to go to a further step to process some data. One of the possible area is images collected from the Robotics developed by AH recently. The project will try to train a model with HPCC Systems Machine Learning (ML) General Neural Network (GNN) which support TensorFlow and apply it new data.

If you are interested in this project, please contact the mentors using the details below.

Completion of this project involves:

  • Producing design documentation

  • Develop GNN ECL code to train a model

  • Checked in code

  • Documentation

Data processing with HPCC Systems involves:

  • Learning HPCC Systems architecture

  • Learning ECL

  • Learning HPCC Systems ML GNN

  • HPCC Systems GNN (TensorFlow)

    • Docker Image

    • VirtualBox

    • Hyper-V (VHD)

    • AWS AMI

Processing image data with HPCC Systems GNN involves:

  • Collecting images

  • Pre-processing the images. For example should we need down size the image?

  • Spraying the image data to thor

  • Use GNN to train the model. There are lots of works needed here. Mainly find suitable hyper-parameters

  • Saving the model

  • Using the model for the new image data.

By the mid term review we would expect you to have:

  • Sprayed image data to Thor

  • Written initial ECL code

  • Trained a model with HPCC Systems GNN with some initial result

Mentor

David De Hilster
Contact Details

Backup Mentor: Xiaoming Wang
Contact Details

Skills needed

  • Git, CMake

  • Build ROS package

  • HPCC Systems Platform and ECL

  • Python, Unix Bash,

  • Machine Learning, Neural Networks, particularly Convolution Neural Networks (CNN)

  • Kera, TensorFlow

  • Ability to build and test the HPCC system (guidance will be provided).

  • Ability to write test code. Knowledge of ECL is not a requirement since it should be possible to re-use existing code with minimal changes for this purpose. Links are provided below to our ECL training documentation and online courses should you wish to become familiar with the ECL  language.

Deliverables

Midterm

  • Sprayed image data to Thor

  • Written initial ECL code

  • Trained a model with HPCC Systems GNN with some initial result

End of project

  • Tuning hyperparameters to improve model training.

  • Documentation

Other resources

All pages in this wiki are subject to our site usage guidelines.