09 October 2019

We are far from being able to coexist with flying robots in our urban landscapes, but one of the demonstrations this year within the WASP arena for public safety was a drone that politely gets out of the way when a person walks towards it.

Olov Andersson
Olov Andersson moving around the drone. Photographer: Thor Balkhed
A drone is flying among the trees in the park at Gränsö Manor, keeping a few metres above the ground and at a reasonable distance from trunks and hanging branches. Olov Andersson, doctoral student within AI and Integrated Computer Systems Division at IDA, crosses the drone’s route as he walks through the park, and the drone politely gets out of the way.

“I’m going to frighten it a bit”, says Olov Andersson, and runs rapidly towards the drone, which backs away just as rapidly.

Olov Andersson is a WASP doctoral student in the Department of Computer and Information Science, at LiU, working to create safe coexistence between autonomous mobile robots and people.

Challenging AI research

Patrick Doherty Photo credit Thor Balkhed“It shouldn’t be necessary for people to keep track of whether drones are flying around in town, or over their heads: the system itself should keep track of where people are and it must be absolutely safe”, says Patrick Doherty, professor in AI at LiU and Olov Andersson’s PhD supervisor.

“This is a highly challenging and difficult AI research topic, and much remains to be done before we can coexist with mobile robots in everyday life”, Patrick Doherty concludes.

Olov Andersson continues the demonstration, and now the drone is not only keeping tabs on tree trunks, lampposts, and the Olov Andersson Photo credit Thor Balkhedunpredictable Olov: it is also collecting large amounts of data that will eventually enable it to reduce its safety distance and gradually fly increasingly closer to obstacles.
“The system learns as it flies, and the safety level has been set at 99.9%”, says Olov Andersson and points at graphs on the computer screen.

“What would happen if we all approached from different directions”, asks Gunnar Bark, responsible for the WASP arenas.

“At the moment, given the computing capacity we have available here, it would probably just give up”, laughs Olov Andersson.

WARA PS demonstration

Around 100 people have gathered at Gränsö Manor in Västervik, all of them working in Sweden’s largest research programme WASP, the Wallenberg AI Autonomous Systems and Software Program. The topic being considered is autonomous rescue services, on land and at sea, with lectures, exchange of experiences, workshops and demonstrations.

“The aim is to demonstrate in practice what we can achieve when researchers from several universities and disciplines work together and with the industrial partners involved in WASP”, says Jesper Tordenlid, from Saab Combitech and responsible for WARA PS, the WASP arena for public safety.

The main demonstration is being held on what is today a rather windy bay, and at first view appears to be similar to the demonstration held last year. More units are involved, and the software and cloud-based services have been updated. On shore, representatives from the WASP board and management watch the manoeuvres, together with doctoral WASP Gränsö 20190516 Photo credit Magnus Johanssonstudents, master’s students, supervisors from the universities involved, representatives for the companies, the Swedish Sea Rescue Society, and the Fire and Rescue Service in Västervik.

An alarm comes in. A high-speed rescue delta wing, from the Swedish Sea Rescue Society and Airpelago, is dispatched to check whether an accident has occurred, and if so, where. It reports back two points of interest to the command centre and two drones fly off. They allocate the search area between themselves completely autonomously, looking for people in the sea. Drone flying at Gränsö Photo credit Thor BalkhedData are sent to the command centre, which deploys a fast boat, a Piraya, that determines the location even more accurately. Now the rescue vessel can travel to the exact place to rescue those in need.

Facing the real world

If other points of interest have been detected in the search area, the drone and boat continue to investigate them and make sure that all those in need of help receive it rapidly. The scenario only takes a few minutes to play out, and the communication between drones, boats and people takes place completely autonomously.
A drone delivering a firs-aid kit Photo credit Thor BalkhedIf a first-aid kit is needed, another drone buzzes off to deliver its load, nice and carefully, close to those who need it.

Present at this year’s demonstration were scientists and doctoral students from SMaRC, the Swedish Maritime Robotics Centre, the largest Swedish programme for underwater robotics.

“The challenges we are facing are similar, and we work with similar technology. There’s a lot we can learn from each other”, says Ivan Stenius, principal investigator at SMaRC.

The underwater robot found it difficult to cope with seaweed, and the initial demonstration took place in the swimming pool. One of the main intentions of the tests and demonstrations in and around Gränsö Manor is just this: to experience the unexpected, difficulties that students, doctoral students and scientists may not be able to predict from laboratory work and theory. Water splashed on the camera lens, rocking boats, gusts of wind, and tangled seaweed, to give a few examples.

“Welcome to the real world”, was the challenge from Jesper Tordenlid in his introduction.

Translated by George Farrants


Contact

More news about WARA PS

WASP at LiU

Latest news from LiU

LiU receives SEK 232 million from the Swedish Research Council

The Swedish Research Council has now decided on the allocation of research grants in four areas. At Linköping University, the area of natural and engineering sciences is to receive the most funding.

A researcher is working together with a test subject.

Our sense of touch consists of 16 unique types of nerve cells

No less than 16 different types of nerve cells have been identified by scientists in a new study on the human sense of touch. Comparisons between humans, mice and macaques show both similarities and significant differences.

Two men in white lab coats with a computer in a lab.

Improving Alphafold to predict very large proteins

The AI tool Alphafold has been improved so that it can now predict the shape of very large and complex protein structures. Linköping University researchers have also succeeded in integrating experimental data into the tool.