Gallery

Industries such as construction, agriculture, and many more require the ability to position and orient payloads with unique geometries. Utilising a swarm of drones to perform these tasks allows not only for the payload-carrying capacity to be scaled by varying the number of drones, but also enables high precision and control. This video shows two small drones carrying a rigid payload between them. The payload is free to pivot, allowing the drones to not only keep it airborne, but also attain a desired position and orientation. This work was conducted by Connor Williams and Jaap Skinner as a Final Year Research Project in Mechanical and Mechatronics Engineering at the University of Auckland.
Collaboratively Carrying a Payload with Two Drones
Aerial manipulation is greatly expanding the potential of drones, removing workers from potential harm. An example of this is installation of components at height, which normally involves a worker at height driving a screw into a vertical wall. However, despite the prevalence of such work in construction and engineering, the challenges of using a drone to drive a horizontal screw have not been extensively explored. Here, an overactuated drone to apply a torque to a peg in a hole. Note the arm on the receptor turning as the torque is applied. This replicates the loading of a screw, as well as the motion constraints applied to the UAV. This work was conducted by Louis Lee and Dulina Senevirathna as a Final Year Research Project in Mechanical and Mechatronics Engineering at the University of Auckland.
Screwdriving UAV Demonstration
A major focus in our drone development is making aircraft which can perform tasks involving physical contact. Both the aircraft and controller must be capable of accepting the forces resulting from the contact, and maintaining position in the face of disturbances.<br />An important aspect is the angle at which the contact is made. In this video, an overactuated octorotor makes and holds contact with a bar (representing a powerline) at five different angles.<br />This work was conducted by Chantelle Singh as part of her Masters in Mechatronics Engineering at the University of Auckland.
Prolonged Physical Contact
Multirotor drones can be required to operate in challenging environments, such as high wind conditions. Rather than attempting to hover, one option is to perch on the object. This requires an adaptable leg, which can grip any potential target branch or rail.<br />This video shows one such leg, which was developed by Junwoo Park in 2024 as part of a summer research scholarship.
Adaptable Leg for Multirotor UAVs
Multirotor UAVs suffer from endurance issues due to limited battery capacity, greatly limiting their use for any task such as surveillance. To remedy this, a system was developed for perching of multirotor UAVs onto agricultural farm posts. The project consists of two major sections: design and testing of a perching mechanism and design and testing of a Computer Vision based system for automated landing, with the aim of the project to automate the perching. Here we see a passive gripper under test in the lab. As the weight of the UAV is applied to the post, the legs are forced inwards, gripping the aircraft securely to its perch. This work was conducted by Tzu-Jui Lin and David Long as a Final Year Research Project in Mechanical and Mechatronics Engineering at the University of Auckland.
Farm Post Perching UAV
The aim of the multi-year Skydiving UAV is to provide a photography platform for skydiving operators.. A major aspect implemented is an autonomous parachute recovery system, to enable safe and reliable recovery. This video shows several phases of flight-testing the parachute recovery system. The UAV is first dropped from a quadcopter in a low-altitude experiment, before being taken up in a Cessna 172M and dropped from an altitude of 750 m. This work was conducted by Lawrence Pascual and Reuben Schuitemaker as a Final Year Research Project in Mechanical and Mechatronics Engineering at the University of Auckland.
Skydiving Photography Drone and Autonomous Parachute Recovery System
Autonomous object avoidance is critical to safe flight. Constructing a flight path from following an object is a means of attaining this without additional proximity sensing, as it is assumed that the object being followed has not collided with anything. This project developed a system for an object to be autonomously tracked by a UAV, replicating its path. In this video, the path of a ball is autonomously replicated by a quadcopter flying around an orchard. This work was conducted by Jay Mills and Blair Eastwood as a Final Year Research Project in Mechanical and Mechatronics Engineering at the University of Auckland.
UAV with On-Board Object Path Following
While much of our work focuses on accuracy of physical interaction, for aerial surveying, such as LiDAR scanning of plantation forests, it is essential to not only maintain an accurate path, but also a consistent speed.<br /><br />In this video, a quadcopter attempts to maintain a consistent path and speed using a Model Predictive Controller (MPC). The path is a circle, inclined at 6 degrees, which is flown at 4 m/s. Given the size of the lab, this is probably the fastest we've flown within it! This work was conducted by Tzu-Jui Lin in 2022 as part of his Doctoral Studies at the University of Auckland.
High-Speed Path Tracking
Normally, a flight path is designed for the aircraft to follow before flight. However, sometimes the aircraft will encounter unexpected obstacles, which it must avoid. This leads to autonomous navigation, where the aircraft sees and avoids an obstacle. Before flying autonomously in a plantation forest (see our related video: https://www.youtube.com/watch?v=99vuafVdsUM), the autonomous path recalculation was demonstrated in the lab. The aircraft is instructed to fly through an obstacle (the box), but instead works its way around, constructing a map with distances to objects to ensure that it will safely avoid collisions. This work was conducted by Tzu-Jui Lin in 2022 as part of his Doctoral Studies at the University of Auckland.
Autonomous Laboratory Navigation
If a drone needs to examine a tree, branches can block sensors and tangle in rotors. To avoid this hazard, the drone needs to be able to push away such obstacles. Here we see a branch-spreading system in action, pushing tree branches apart to peer through below the canopy. This is an irregular, unstructured object interaction, in contrast to the initial tests with a controlled load.<br /><br />This work was conducted by Shahab Kazemi and Joao Buzzatto, based on the PhD Research of Pedro Mendes.<br /><br />Also take a look at the initial tests, which used a known, structured load to put the system through its paces.<br />https://www.youtube.com/watch?v=12NPzVsAh68&list=PLsOGHpYv1BDByIexQ0je5oZ9lC6wQhwth<br />https://www.youtube.com/watch?v=iyBWLXAwF8g&list=PLsOGHpYv1BDByIexQ0je5oZ9lC6wQhwth&index=2
Separating Tree Branches with a Drone
Conventional drones are under-actuated: the only way to produce forces other than vertical is to rotate the entire aircraft. The most apparent consequence being the need to tilt to move horizontally, which makes precise positioning difficult. Fully- and over-actuated drones are able to produce forces in other directions without tilting the aircraft.<br /><br />Here we see an over-actuated drone using fixed-angle tilted rotors to move horizontally without the need to rotate the entire aircraft. This particular drone also differs in being "heterogeneous", as the motors and propellers are not all the same: most drones, whether under- or over-actuated, are "homogeneous", using the same motors and propellers across the aircraft.<br /><br />This work was conducted by Research Assistant David Yang, and is based on the PhD Research of Salim Al-Zubaidi.
Horizontal Thrust Demonstration on an Over-actuated Drone
Physical human drone interaction (PHDI) with underactuated drones has been widely explored due to its diverse applications from providing guidance to the visually impaired to virtual reality gaming. An opportunity exists in exploiting the decoupled translational and rotational dynamics of over-actuated drones for PHDI. ​Watch as an overactuated aircraft gets dragged through the air during human interaction, the aircraft accepting the drag disturbance while maintaining controlled flight. This work was conducted by Iris Liu and Suvarna Swaraj as a Final Year Research Project in Mechanical and Mechatronics Engineering at the University of Auckland.
Human Drone Physical Interaction
Load More… Subscribe