To achieve AOS’ mission of developing an Australian off-road autonomous capability, field testing is essential, and this includes testing the ability of the vehicles to safely detect, identify and avoid people walking, sitting or lying down, children and animals – both our wide range of native ones, as well as farm animals. Currently, the majority of obstacles used for this purpose are either static or have simple, easily predictable movements, especially when considering off-road obstacles. To obtain meaningful data regarding the performance of the sensing and path planning algorithms of autonomous vehicles, we require sophisticated obstacles that mimic practical situations encountered by the vehicles as closely as possible, without endangering people or animals.
To achieve this, AOS has developed the Pug autonomous pedestrian. With a custom-developed chassis and software modified from AOS’ Kelpie autonomous land vehicle, the Pug allows for emulating pedestrians and animals in a safe and controlled manner. It is able to be tasked with multiple behavioural algorithms, resulting in smart and natural decision making. Thanks to the Pug’s design, it can be used on traditional roads and concrete, as well as off-road and on rough terrain.
The Pug’s introduction offers a new capability to support the testing of autonomous vehicles in Australia. AOS is looking forward to launching the Pug in May, and available for autonomous vehicle testing around Australia.
On the left: A picture taken from the Kelpie of the static pedestrian mannequin to be replaced with the smart, Pug-powered mannequin. On the right: Bryan working on the first Pug prototype.