Robots to ResQu our rainforests

Unmanned autonomous aircraft are being trialled in surveys to detect weeds in difficult rainforest terrain. This research is part of Project ResQu, established to develop the use of unmanned autonomous vehicles for search and rescue operations and weed surveillance programs.

The Challenge

Protecting rainforest from weeds

Our rainforests are precious – less than two per cent of Australia's land area, confined to small patches clustered mostly in inaccessible, mountainous regions along the tropical coast. It's important we look after these precious habitats. Unfortunately, a purple-leaved weed, Miconia calvescens, has invaded some areas of rainforest. Miconia is an unusually aggressive invader and has potential to cause irreversible damage to native plant and wildlife populations.

Our Response

Unmanned aircraft search for weeds

Project ResQu is a two year, $7M project led by the Australian Research Centre for Aerospace Automation (ARCAA) in a collaborative project between the Queensland University of Technology (QUT), CSIRO, Boeing and Insitu Pacific with the support of the Queensland State Government Department of Science, Information Technology, Innovation and the Arts.

Developed by robotics researchers at CSIRO in partnership with Biosecurity Queensland, the unmanned helicopters locate weeds using sophisticated imaging technology. They offer a safer, more convenient way of mapping weeds in remote and difficult terrain.

Unmanned Aerial Vehicles (UAVs) give weed spotters a safer job, allowing them to review footage at their leisure rather than take to the skies and eyeball weeds on the go from full-sized manned helicopters, which is expensive and can be dangerous. As the information is stored, it allows different departments to share data for comparison over time.

Robots to ResQu our rainforests

Show transcript

Transcript

[Music plays and text appears: Robots to ResQu our rainforests]

[Image changes to show various pictures of a rainforest]

[Image changes to show Doctor Gary Fitt – Science Director, Biosecurity Flagship, CSIRO]

Dr. Gary Fitt: Miconia is an invasive tree, and it’s one of the world’s top 100 invasive plants, so it’s something to be really concerned about.

[Image changes to show various pictures of Miconia trees]

It’s a plant that’s invaded rainforests in many other countries. You look at Tahiti, 25% of their rainforest is now Miconia.

[Image changes back to Doctor Gary Fitt]

Something we don’t want in Australia.

[Image changes to show a helicopter on the ground]

[Image changes to show men in a rainforest]

[Image changes to aerial pictures of a rainforest from a helicopter]

[Image changes to show Mick Jeffery – Project Coordinator (Tropical Weed Eradication), Biosecurity Queensland]

Mick Jeffery: Most of the work we do is on ground. We’ve got teams that bash through the rainforest, searching.

[Image changes to show various pictures of people searching through the rainforest]

But we don’t see everything from the ground; we also need to detect things from the air.

[Image changes to show a helicopter on the ground]

[Image changes to show a group of people sitting in a helicopter]

But flying in a helicopter with our staff and spotters in there, it can be dangerous, they’re relatively close.

[Image changes to an aerial shot of a rainforest from a helicopter]

[Image changes back to Mick Jeffery]

In the long term we would hope to take that human risk element out of aerial operations.

[Image changes to show a rainforest covered in fog]

[Image changes to show a rainforest]

[Image changes to show remote controlled helicopters being operated by a group of people in the field]

[Image changes back to Doctor Gary Fitt]

Dr. Gary Fitt: The real challenge with an invasive plant like Miconia that’s invading a very difficult environment, a difficult terrain, is to be able to find those plants early enough to make a difference.

[Image changes to show various pictures of Miconia trees]

[Image changes to show people setting up remote controlled helicopters in the field]

So it’s really a needle in a haystack problem, and you need to be able to get smart technology to the right places, and that’s where the autonomous platforms that we’re developing at CSIRO really play a big role.

[Image changes to show a group of people working at a computer]

[Image changes to show Rowland Marshall – Project Manager, Queensland University of Technology]

Rowland Marshall: Project ResQu is a $7 million two year project that aims to fast track the utilisation of unmanned aircraft in the civilian environment.

[Image changes to show the camera on the autonomous helicopter being operated]

[Image changes to show an autonomous helicopter taking off]

What we’ve done today is demonstrated utilisation of an unmanned aircraft in the detection of Miconia weed.

[Image changes to show an autonomous helicopter hovering above the ground]

We are performing this research project with CSIRO, Queensland University of Technology, Boeing, Insitu Pacific, and the Queensland State Government.

[Image changes to show a man working at a computer]

[Image changes back to Rowland Marshall]

In order to complete the work we’ve also been liaising with Biosecurity Queensland.

[Image changes to show an autonomous helicopter landing on the ground]

[Image changes to show Doctor Torsten Merz – Project Leader, Project ResQu – Miconia Helicopter, CSIRO]

Dr. Torsten Merz: This is the Miconia helicopter prototype system. Yeah, we just completed the last of our demonstration flights.

[Image changes to show Doctor Torsten Merz demonstrating the camera operation on the remote controlled helicopter]

So during the cruise flight it will sweep like that and it takes five images per second.

[Image changes to show Doctor Torsten Merz pointing to an area behind him]

We flew up into this area over there, where last week they found Miconia during a manned helicopter survey.

[Image changes to show an aerial view from the autonomous helicopter]

Tricky part with these missions is we fly relatively far away from our takeoff location, so the helicopter is not a very big aircraft, so once it goes beyond a certain distance it’s just a little dot in the sky.

[Image changes to show the autonomous helicopter flying in the air]

And on top of that we fly at low altitude, very close to the rainforest canopy.

[Image changes to show a man viewing the flight path of the autonomous helicopter on a computer]

[Image changes to show Doctor Torsten Merz and another man working on the autonomous helicopter]

The radar is used for detecting terrain obstacles which are not in our maps.

[Image changes to show an aerial view from the autonomous helicopter]

You need a good situational awareness to conduct these flights safely without human interaction.

[Image changes to show a group of people working at a computer]

[Image changes to show various computer images of rainforest]

A manned helicopter survey was done in the same area where we conducted our unmanned helicopter survey, so we had a quick analysis of the images we took during the surveys and we spotted some Miconia during that analysis which hasn’t been spotted during the manned survey.

[Image changes to show a man pointing to a Miconia plant on a computer image]

[Image changes back to Doctor Torsten Merz]

This is a great outcome.

[Image changes to show an autonomous helicopter hovering above the ground]

[Image changes to show two men working at a computer]

Dr. Gary Fitt: I guess in the biosecurity space effective surveillance is a key, we need to be able to detect incursions quickly and accurately, and when they’re present at very low densities.

[Image changes to show men working in the field]

[Image changes to show an aerial view from the autonomous helicopter]

[Image changes to show computer image of a rainforest]

[Image changes to show an autonomous helicopter hovering above the ground]

And so we need to have technologies like the autonomous helicopter, or other autonomous platforms, that can work 24/7 detecting incursions.

[Image changes to show an autonomous helicopter landing on the ground]

[Image changes back to Rowland Marshall]

Rowland Marshall: Moving forward from the project we’d really like to see the Miconia system be accepted by a commercial partner and move on to full development and deployment.

[Image changes to show an autonomous helicopter on the ground]

Project ResQu’s been a really successful project and I think it’s been able to show the viability of unmanned aircraft.

[Image changes back to Rowland Marshall]

We’re really excited about the future and being able to see them benefit people more and more.

[Image changes to show a man working at a computer]

[Image changes to show a computer image of a topographic map]

[Image changes to show a computer image of a rainforest]

[Image changes back to Doctor Gary Fitt]

Dr. Gary Fitt: As we become more and more threatened by invasive species we have to use the best technology we find, and autonomous systems from our point of view, fitted with appropriate sensors, is really the way to go.

[Image changes to show an autonomous helicopter taking off]

[Image changes to show various logos and text appears: Project ResQu is conducted with the support of our sponsors CSIRO, QUT, Queensland Government, Boeing, Insitu Pacific, ResQu

[Music plays and CSIRO logo appears with text: Big ideas start here www.csiro.au]

Hide transcript

The Results

More weeds recorded than by eye

CSIRO built the prototype UAVs over two years and trialled them in difficult conditions. They performed better than expected, finding Miconia plants in the rainforest that hadn't been spotted before. They're proving a safe and convenient alternative for traditional aerial weed surveys.

Dr Torsten Merz preparing the unmanned autonomous helicopter.

The unmanned Project ResQu helicopters were deployed to survey rainforests at El Arish, near Cairns during August 2014.

They found not just Miconia but several other weed species. The robotic helicopters can navigate obstacles without human control while recording locations and images for biosecurity staff to scan for evidence of weeds.

The next step, pending extra funding, will be to fully automate weed identification and further improve UAV dependability and safety.

Enquiries

Have an enquiry about this page?

Contact us

Do business with us to help your organisation thrive

We partner with small and large companies, government and industry in Australia and around the world.

Contact us now to start doing business