Technology facilitates ‘breakthrough age’ in citizen science

Lindy Orwin pulls a long-horned beetle in a petri dish out of her handbag and plonks it on the table in front of me.

Clipped to my phone is one essential tool in the citizen scientist’s armoury – a macro lens which costs less than $30, but which is powerful enough to reveal the insect’s spindly legs, dramatic antenna and other finer anatomical details.

According to Lindy’s husband Randy, who is running a ‘Tech in the Field’ workshop ahead of this year’s Australian Citizen Science Association conference, such details could prove crucial for identification purposes.

If I’d stumbled across this beetle in the wild, with a few taps of a button, I could upload images to something like iNaturalist, a crowdsourced species identification and organism recording tool which has so far collected more than 165 million observations.

“(iNaturalist) is not just something that you upload your photos to,” says Randy, who is Assistant Director for Learning Technologies at the University of Washington Information School.

“(It) is a community where people who are incredibly passionate about a single species have a search setup, (so) that anytime you upload something that happens to match the family or the genus of that species, they get a notification.

“And these people … they log on immediately, and they go try and identify those things.”

It doesn’t work all the time. Recently, for instance, he and Lindy took their underwater drone into waters near their home in Queensland.

Citizen science still has an undeserved image problem among the broader science community.

Andy Ridley

“I found this really strange critter down the bottom of Tin Can Bay, and I’m like, ‘What in the world is that thing?’” Randy explains.

Even after uploading video and stills to iNaturalist and one of the marine Australia ID facebook groups, they drew a blank.

“So I did a whole bunch of research and came to the conclusion that what I was looking at was what’s called an ‘upside down jellyfish’. I had never even heard of (it),” Randy says.

The practice of citizen science has exploded in recent years, with iNaturalist alone doubling its number of observations and participants each year since 2012. Other popular nature-based platforms include eBird, FrogID and Fungimap.

But citizen science still has an undeserved image problem among the broader science community, says Andy Ridley, founding CEO at the Cairns-based Citizens of the Great Barrier Reef and former leader of the global movement Earth Hour.

“Citizen science is still not recognized in a way that I think it should be,” Ridley told more than 150 delegates at the ACSA conference.

“My passionate belief – and I’ve seen it with Earth Hour – is you could mobilise billions of people if you wanted to, but we need to be taken more seriously as an innovation, rather than a kind of ‘Oh, that’s nice’.”

One of the organisation’s key projects is the annual Great Reef Census, which I participated in two years ago.

The Census involves snorkellers, divers and assorted other citizen scientists jumping in the water with GoPros and taking images of individual coral reefs.

The snapshots are later analysed by a team of 6500 citizen scientists from more than 60 countries, with the information used to help inform decision-making about where to prioritise resources and recovery efforts on the reef.

To date, although 105,963 survey images have been collected, less than one-fifth (17%) of the reefs in the Great Barrier Reef have been surveyed.

But working with citizen scientists is not without its challenges, Ridley concedes.

They can be gung-ho about taking survey photos when in the water, but run out of steam when it comes time to do the grunt work.

“One of the biggest challenges if you’re trying to get photographs of the Great Barrier Reef – your dive instructors, or your tourists, or your skippers, or whoever it is – is that trying to get the images out of the cameras is really hard work,” he explains.

“People might get the pictures but then when they get home, they go to the pub, or they go and see their family, and you never see the pictures.”

(I feel a twinge of guilt when I recall heading for sunset cocktails at the bar, rather than my laptop.)

[Citizen science is] at the beginning of a breakthrough age that could revolutionise what we can do.

Andy Ridley

To overcome this all-too-human hurdle, Ridley’s team partnered with Dell Technologies, Intel and the University of Queensland to come up with ‘edge devices’ which are rugged enough to withstand heat and salt water and can be deployed on lots of different boats.

When citizen scientist surveyors come out of the water, they can upload what they’ve captured straight onto an edge device.

“(The boat) might be somewhere very remote, but when it comes back into range, it will automatically upload (the images),” Ridley explains.

When it comes to analysing the uploaded images, Citizens of the Great Barrier Reef has also added AI-powered tools to help categorise the coral in individual images in seconds rather than seven-plus minutes – with high levels of accuracy.

“We want the science to be rigorous,” Ridley says.

With the use of these and other technologies, he adds, citizen science was “at the beginning of a breakthrough age that could revolutionise what we can do”.

But among citizen science criticisms are that ‘amateur’ observations lack rigour, with different observers having varied skills in species identification, which could affect the reliability of their data.

A paper published earlier this year in Citizen Science: Theory and Practice noted that while citizen science could generate large volumes of data, “much of the reluctance associated with the use of citizen science data within the scientific community surrounds uncertainty about its accuracy”.

However, Randy Orwin challenges this notion, pointing to a paper published in People and Nature which revealed that across all free automated plant identification applications, 85% of images were identified correctly in the top five suggestions, and 69% were correct with the first suggestion.

“It’s not so much the accuracy of the observations themselves – what they’re finding is (that) the issue is the accuracy of the GPS location, and how inconsistent it is,” Randy says.

For that reason, part of his workshop covered how citizen scientists could work with mobile phones, GPS units and DSLR cameras to improve the positional accuracy of their observations.

The usefulness of observations dramatically increased with geolocation – and provided hard evidence that could help support protected species.

We have good longitudinal data over extended periods of time, because citizen scientists are so passionate.

Randy Orwin

Randy pointed to a recent case where an Illawarra property developer dismissed 20 years of citizen scientist reports of platypus in the local waterways as “unreliable”.

Enraged by the developer’s claim that it “did not detect any platypus or obtain conclusive evidence of their habitation along the length of Macquarie Rivulet within the study area”, one woman used her mobile phone to film one of the animals swimming at the site of the developer’s proposed bridge.

The developer is now taking the Shellharbour Council to the Land and Environment Court (LEC) after its plans to build the bridge and a 227-home housing development were rejected.

“We have good longitudinal data over extended periods of time, because citizen scientists are so passionate,” says Randy.

“They just do it because they love to do it, and they care so much for the environment.”