R2024b

 Offroad Autonomy Support Package: Follow paths with MPPI algorithm (October 2024, Version 24.2)

Use the offroadControllerMPPI System object™ to create a controller that uses the Model Predictive Path Integral (MPPI) algorithm, a derivative-free optimization approach, to compute velocity commands and an optimal trajectory for autonomous offroad vehicles and heavy machinery. The controller follows the reference path generated by a global planner. You can tune the parameters of the controller for better path following, path alignment, obstacle avoidance, and smoothness.

offroadControllerMPPI requires the Robotics System Toolbox™ Offroad Autonomy Library support package. You can download the support package from the Add-On Explorer. For more information, see Install Robotics System Toolbox Offroad Autonomy Library Support Package.

 Offroad Autonomy Support Package: Develop autonomous systems for offroad vehicles and heavy machinery

Develop autonomous systems for offroad vehicles and heavy machinery in agriculture, construction, and mining applications with the Robotics System Toolbox Offroad Autonomy Library. This support package provides specialized tools for designing, simulating, and validating autonomy algorithms. Through integration with Unreal Engine^®, this support package enables photorealistic scenario simulations to test and refine the performance of offroad vehicles, such as dump trucks and backhoes, under diverse conditions. See the Offroad Autonomy for Heavy Machinery category for more information.

You can download the Robotics System Toolbox Offroad Autonomy Library support package from the Add-On Explorer. For more information, see Install Robotics System Toolbox Offroad Autonomy Library Support Package.

This image shows a backhoe, simulated in the Unreal Engine Construction Site Scene, excavating material and loading into a dump truck, with the lidar MATLAB^® simulation next to it.

The Offroad Navigation for Autonomous Haul Trucks in Open Pit Mine (Navigation Toolbox) example now also supports the Simulation 3D Physics Dump Truck block and Offroad Pit Mining Scene for simulating the autonomous pit mine navigation.

This support package also comes with these new examples.

• Simulate Earth Moving with Autonomous Excavator in Construction Site

• Simulate Construction Vehicles in Unreal Engine for Material Handling

• Extract Scene from Lidar Data

• Plan Collision-Free Path for Excavator Arm in MATLAB with Lidar Data

• Survey Pit Mining Site with RTK GPS and Point Clouds

Offroad Autonomy Support Package: Simulate a backhoe vehicle using an Unreal Engine 3D simulation environment

You can now use the Simulation 3D Physics Backhoe block to simulate a backhoe vehicle in a Simulink^® 3D world or Unreal^® environment provided in the Robotics System Toolbox Offroad Autonomy Library.

You can download the Robotics System Toolbox Offroad Autonomy Library support package from the Add-On Explorer. For more information, see Install Robotics System Toolbox Offroad Autonomy Library Support Package.

This image shows the backhoe vehicle in the Offroad Pit Mining Scene.

Offroad Autonomy Support Package: Simulate a backhoe vehicle using an Unreal Engine 3D simulation environment

 Intelligent Bin Picking: Design and simulate an intelligent bin picking system in Simulink

Use this new example series to learn how to design an intelligent bin-picking system in Simulink and simulate it using Unreal Engine environments.

Intelligent Bin Picking System in Simulink®

• Design Bin Picking Scene and Sensors in Unreal Engine®

• Design Camera Perception Component to Detect Bin Items

• Design a Trajectory Planner for a Robotic Manipulator

Intelligent Bin Picking: Design and simulate an intelligent bin picking system in Simulink

 Manipulability Analysis: Find high-manipulability regions in robot manipulator workspaces

Use the generateRobotWorkspace and manipulabilityIndex functions to approximate the reachable space of a robot model end-effector and determine how well the robot can manipulate objects in different regions of the workspace, respectively. Use the showWorkspaceAnalysis function to visualize the workspace, encoded by the manipulability values.

These features enable you to identify regions in the robot workspace where the robot can achieve greater ranges of motion for tasks that require extensive movement, such as bin picking where the end-effector could have highly varied picking approaches based on the orientation and position of the object to be picked. You can also use the manipulability index as a metric for selecting inverse-kinematics solutions.

The first figure shows an alpha shape of the robot workspace in an obstacle environment. The second figure visualizes the workspace in an empty environment by mapping the potential positions of end effectors to the manipulability values of their corresponding joint configurations, depicted through a color gradient.

Manipulability Analysis: Find high-manipulability regions in robot manipulator workspaces

dlCHOMP Object: Generate C and C++ code

The dlCHOMP object and its optimize object function support C and C++ code generation using MATLAB Coder™.

Euler Angle Conversions: Convert to and from other rotations using additional Euler angle sequences

The Coordinate Transformation Conversion block and quat2eul function now support additional Euler sequences. These are all of the supported Euler sequences:

• "ZYX"

• "ZYZ"

• "ZXY"

• "ZXZ"

• "YXY"

• "YZX"

• "YXZ"

• "YZY"

• "XYX"

• "XYZ"

• "XZX"

• "XZY"

Robot Models: Use additional manipulators introduced to robot model library

You can now retrieve these additional manipulator robots from the robot model library using the loadrobot and importrobot functions:

Robot Model	Mesh Visualization
"robotiqEPick4CupVacuumAssembly"
"robotiqEPick2CupVacuumAssembly"
"robotiqEPickVacuumCup"
"robotiqEPickVacuumCup200mm"

Rigid Body Visuals: Scale the size of meshes when adding to a rigid body

Specify an xyz-scale for meshes when adding them to a rigid body using the addVisual object function. You can get the scale information from rigid bodies using the Scale field of the meshData output of the getVisual object function.

RRT Goal Regions: Specify multiple goal space regions

The plan object function of manipulatorRRT now supports specifying multiple workspace goal regions as a cell array using the goalRegion argument. This function samples goal configurations uniformly from the specified goal regions.

Kinematic Motion Models: Calculate derivative of multiple commands and states at once

Calculate the derivative for multiple states and commands at once for a kinematic motion model System object. See the derivative object function for more information.

UR Series Manipulators: Control cobot using data exchange over RTDE protocol

The Robotics System Toolbox Support Package for Universal Robots UR Series Manipulators enables you to control a UR Series cobot using data exchange over the Real-Time Data Exchange (RTDE) protocol. Select RTDE as one of the connectivity options in the Hardware Setup screen, and follow the instructions to set up the required configuration for connecting to either the URSim simulator or the UR Series manipulator hardware.

To connect to a UR Series cobot over RTDE, use the new urRTDEClient object that establishes a TCP/IP connection from MATLAB to the UR controller. Along with the existing functions available in the support package, you can also use these new functions designed for the urRTDEClient object to interact with the cobot (either simulated in URSim simulator or the physical hardware) and perform manipulation tasks:

• sendServoJCommands

• sendSpeedJCommands

• manageGripperAttachment

• executeURScriptCommand

• actuateGripper

New Examples

This release contains these new examples:

Offroad Autonomy Support Package

• Simulate Earth Moving with Autonomous Excavator in Construction Site

• Simulate Construction Vehicles in Unreal Engine for Material Handling

• Extract Scene from Lidar Data

• Plan Collision-Free Path for Excavator Arm in MATLAB with Lidar Data

• Survey Pit Mining Site with RTK GPS and Point Clouds

Intelligent Bin Picking

• Intelligent Bin Picking System in Simulink®

• Design Bin Picking Scene and Sensors in Unreal Engine®

• Design Camera Perception Component to Detect Bin Items

• Design a Trajectory Planner for a Robotic Manipulator

Miscellaneous

• Palletize Boxes Using Cobot with Simulink 3D Animation™

• Estimate Pose of Fixed Camera Relative to Robot Base

• Estimate Pose of Moving Camera Mounted on a Robot (Computer Vision Toolbox)

• Workspace Analysis for Manipulators

• Plan Path in Warehouse Scenario with Unseen Obstacle Avoidance