PX4 Developer Summit 2021

Proper tuning of low level controllers is a critical step when setting up a new drone. Trial and error manual tuning requires good piloting skills, tuning experience and is often only possible for PID controllers. In this presentation we will see how this process can be automated using an online (i.e.: running in real-time on the flight controller) self-tuning controller based on indirect adaptive control theory that can be used on various types of drones in order to tune the inner control loops at a click of a button.

Although we’ve heard of the term ‘Zero Trust’ for a while now, the concept has recently picked up momentum with businesses looking to proactively protect their data and infrastructure. With the shift to the cloud, Zero Trust is now almost a preferred business philosophy for CIOs and Chief Information Security Officers (CISOs), who are tasked with protecting their systems from outside attacks as well as from within the organization.

Ignition gazebo is a opensource robotics simulator, which provides high fidelity physics, rendering and various sensor models as part of the Ignition Robotics project. Ignition Gazebo is derived from the popular Gazebo simulator, which has been one of the most popular simulators in robotics for the academia and the industry. Since ignition gazebo provides interfaces for advanced rendering capabilities and sensors, testing simulated vehicles in ignition gazebo provides a flexible environment for testing advanced capabilities of the vehicle. In this talk, we discuss the new integration of the PX4 simulation environment into Ignition Gazebo and explain how ignition gazebo can help develop robust autonomous vehicles.

Unsecure drone systems may pose threats to our privacy, cyber security and our physical safety. Secure Systems Research Centre at Technology Innovation Institute (TII/SSRC) wants to be involved in making the current drone platform better, more secure and resilient for the safe use of hobbyists, commercial and governmental authorities. TII/SSRC will now be presenting how the PX4 Firmware is used to enable drone security research and what basic security improvements are being developed for encryption, memory protection and a more secure hardware. In addition, TII/SSRC will be presenting a security enabled hardware platform onto which the PX4 Autopilot software was ported and with those changes extreme levels of performance are achieved. The hardware platform is based on RISC-V application grade CPU cores, and it is capable of running multiple operating systems in parallel. Furthermore, the hardware platform integrates hardware accelerated cryptography on a single SoC. Porting the PX4 Autopilot on this new HW is done in order to enable new exploratory paths in drone software development for TII/SSRC's research partners.

It has long been a challenge in many fields to measure atmospheric parameters in free space. These can include measurements of solid particles, radiation, gases, contamination and others. Today’s solutions often lack flexibility, UAV’s are mostly not using real-time data to adjust the monitoring process. They cannot operate under extreme weather or require highly qualified personnel. This panel discussion debates platform-independent solutions for the above-mentioned challenges and brings specific examples of how the solutions were implemented and used to solve problems in the real world. It also addresses the importance of PX4 and its contribution in creating similar projects.

NXP mobile robotics team makes available opensource industrial and automotive designs focused on PX4 and Mobile Robotics encompassing SmartBMS with CAN, UAVCAN Nodes, CAN network switches, T1 2-wire Ethernet, Ultrawideband Localization,  Vision and AI accelerated Linux with ROS2, Flight controllers and more. Learn about existing and upcoming development boards and opensource software enablement which can be leveraged build your own hardware or develop software for PX4, UAVCAN, and ROS2.

While the task of integrating PX4 into the DDS world has been dealt with by means of the micro-RTPS bridge so far, the present proposal hinges on the need for a more mature, flexible and secure interface with DDS that would be carried out by, first, micro XRCE-DDS, an then by micro-ROS, and can be seen as a natural extension of the work done so far over the micro-RTPS bridge. micro-ROS is a Robot Operating System specifically tailored for embedded and resource-constrained platforms, such as MCUs. It promotes these to the ROS 2 world, and uses as a middleware eProsima Micro XRCE-DDS, that is, DDS for eXtremely Resource-Constrained Environments, a software solution allowing the integration of microcontrollers into the DDS world. As an extension of the last talk at last year's summit, where Nuno and Jaime (eProsima’s CEO) presented a general overview of what would be required to do the migration from the micro-RTPS bridge to micro-ROS, this time, we will now go in-depth on the core concepts and technical details behind the migration from the current micro-RTPS implementation to having a micro-ROS capable PX4 ecosystem. Note: We request that, in case of acceptance, that this call is scheduled to the 15th of September, by impossibility of presenting it Tuesday the 14th.

This session is aimed at providing updates to the ongoing efforts between 96Boards, Gumstix and PX4 to create Heterogeneous Compute Platform for Drones using the 96Boards SoM spec.
This session will go into detail on Why such a platform is required, What are the use-cases and how the collaboration between 96Boards, Gumstix and PX4 brought the project to life.

The Camera Auto Mount Pivoting Oblique Survey (CAMPOS) is a new method for increasing the efficiency while performing UAS surveys, it can either increase the area a vehicle is able to cover in a single flight or reduce the flight time needed to map a given area when compared to traditional, nadir only surveys. This allows for doubling the efficiency in 2.5D mapping missions or even greater factors for surveys targeting the creation of 3D models. CAMPOS is essentially a software feature that intelligently changes the camera orientation between shots along a flight traverse in a surveying mission, allowing for an increased horizontal field of view, resulting in smaller flight paths through less lines. This means even existing UASs can benefit from this feature through a software update, the only requisite is that the camera mount is actuated on the roll axis, which is also an easy mod for fixed-wing vehicles. This work was peer reviewed and published at ISPRS XXIV Congress. Igor hopes to showcase this new feature that is a result of contributions to the PX4 and ArduPilot flight stacks, MAVLink protocol and the QGC ground control station. Hopefully this will increase the adoption of this valuable tool that became available before similar solutions from big players in the industry.

Currently, CubeSats don't have much wider use of Operating Systems (OS) which, in theory, could reduce the time taken for the development of CubeSat. If there would have been a CubeSat OS, the success rate of the University CubeSat Mission would increase, as there would be more time available to test the hardware to increase the mission success rate. Antarikchya Pratisthan Nepal, is trying to develop such software that would be easy to use while reducing the time it takes to design CubeSats. For this first step, PX4 Autopilot is being integrated as a mission, which is to be placed as one of the payloads aboard the 3U CubeSat and launched. This mission is to test how the PX4 and its constituents behave in a radiation-prone, vacuum space environment. The initial test hardware is based on the Pixracer v1.0 flight controller board, which was also used in past CanSat training here in Nepal. After a successful Initial test and once enough data are collected the team will be moving forward with developing an airframe similar to is done for drones. This allows for the utilization of COTS drone components to design and fabricate a CubeSat while reducing the complexity and allowing more time for extensive testing of the system.

PX4 powered unmanned autogyros offer new possibilities in the field of aerial work, thus filling the existing gap in the offer of unmanned drones. In the lecture you’ll learn how the open-source TF-G2 autogyro was designed and created, what problems we faced during its development and what is the current state of progress. TF-G2 autogyro is primarily a platform for operator training. However, thanks to its smart design it is able to carry useful payload and perform measurements in the air. An interesting feature of an autogyro-type drone is its resistance to worsened weather conditions. Most parts of the drone are 3D printable, leading to its easy repairability.