Robotics & Autonomous Systems Technical Talks

In this session, we will introduce the workflows for connecting MATLAB/Simulink and ROS using the ROS Toolbox^™. You will learn software features, templates, and best practices to help you implement common ROS programming constructs in MATLAB and Simulink. You will also learn how to combine MATLAB, Simulink, and State flow as modeling tools for different types of algorithms within a Simulink block diagram.

In the end, we will show a demonstration of the concepts above through an autonomous object tracking example. This example implements best practices with MATLAB and Robotics System Toolbox. It also highlights the modularity of MATLAB and ROS by showing the algorithm using real and simulated TurtleBot^® robotic platforms, as well as a webcam. Two new products were introduced in 2 to complement the capabilities of Robotics System Toolbox: Navigation Toolbox and ROS Toolbox.

MATLAB and Simulink support a live connection with ROS network for your ROS based desktop simulation. In addition, you can generate stand-alone C++ nodes from MATLAB and Simulink through automated code generation. Starting with R2021b, it also supports embedding a deep learning network into ROS nodes using GPU coder (e.g., generate CUDA code of object detection algorithm from MATLAB Script using GPU Coder). 

This webinar discusses automated C++ ROS / ROS2 node generation from MATLAB scripts and Simulink models.  We will also highlight ROS Toolbox feature updates and propose the workflow that successful users have adopted.

Highlights

- ROS and ROS 2 node generation and deployment from MATLAB scripts

- ROS and ROS 2 node generation and deployment from Simulink models

- ROS-compatible CUDA node generation from Simulink using GPU coder

- Working with generated ROS nodes

Increasing automation in robotics leads to a rise in applications of advanced AI models. Machine learning and Deep learning are finding more and more uses in Robotics in areas such as sensor fusion, perception, planning and control. Developing robust and accurate deep learning models that need to make precise and fast decisions are becoming increasingly more important in the industry. This webinar will highlight the tools in MATLAB used for developing deep learning models in use for robotics. We will go through topics such as an GUI-based workflow for labelling LiDAR or video data with automated labelling. Repurposing pre-built models with minimal coding in images. Developing Reinforcement Learning models for autonomous robots. And deploying MATLAB code into hardware or simulations.

Highlights:

- GUI based workflow for Labelling Data

- Developing models based on pre-trained models

- Maximizing efficiency using GPU accelerated hardware

- Deployment of trained model for simulation and prototyping

AI is transforming engineering in nearly every industry and application area. Smart factories leverage AI for various applications such as predictive maintenance and quality inspection. Identifying product defects and reducing manufacturing errors can help reduce labor and manufacturing costs significantly.

Highlights:

- Trending towards Smart Factories

- Collaborative opportunities

- Model Based Design for autonomous systems development

Unmanned Aerial Vehicle (UAV) usage is increasing across a variety of commercial applications such as delivery, agriculture, construction, and others.  Many UAV applications incorporate technologies for autonomous operation and developing autonomous UAV systems and applications require multi-domain technologies and need simulation to ensure proper operations prior to testing on real hardware to reduce risk and rework. Model-Based Design provides an integrated workflow that allow engineers and researchers to develop UAV systems and applications, evaluate through virtual testing, and deploy to UAV hardware. 

Highlights:

- Modeling methodologies for UAV systems

- Designing autonomous UAV algorithms

- Simulating autonomous UAV applications with sensor models

- Automating verification and validation tasks to ensure system robustness

- Automatically generating software and deploying to autopilot and onboard computer hardware

- Analyzing post-flight telemetry data

Autonomous systems continue to be developed and deployed at an accelerated pace across a wide range of applications including automated driving, robotics, and UAVs. The four main components of an autonomous system are sensing, perception, decision making, and action execution and control.

In this webinar, we will focus on the perception component of an autonomous system. We will demonstrate a sensor fusion development framework that includes an integrated test bed, algorithms, and metrics for performance analysis. We will look at the strengths and weaknesses of different sensor modalities for localization and multi-object tracking. We will also explore how measurements from different sensor modalities such as radar and lidar can be combined to improve the quality of the perception system.

Multiple examples will be used to highlight concepts that are critical to architecting and integrating systems that perform sensor fusion.

Highlights

You will be able to learn:

- Why sensor fusion and tracking are a key part of perception systems

- How to combine measurements that are taken from different sensor to improve pose estimates and situational awareness

- How to explore system level trade-offs including sensor accuracy, location, and update rates 

Robotics deals with modelling and controlling mobile robots with the aim of high precision implementation. This is the reason why designing and navigational algorithm development, early testing, and target hardware deployment often than not becomes a challenge.

In this session, will be sharing ways to address this by showing how MATLAB and Simulink can help you mitigate object detection, data processing, and control creation to implement an integrated AMR design workflow.

Highlights:

- Generating occupancy maps and simulation environment from images

- Leveraging out-of-the-box tools to rapidly prototype navigation algorithms

- Integrating external simulators by utilizing a ROS interface