Go Jauntly applies AI to seek scale via ‘greener’ walking routes

Unless you’ve been on very extended digital detox, you’ll have noticed algorithms don’t exactly have the greatest reputation these days, saddled as they are with pervasive questions of bias and inequity. Not to mention the dubious content amplification choices of social media platforms.

Nor has 2020 helped their automated cause — with, in just one example, outraged UK students leading chants of ‘fuck the algorithm‘ this summer as they were assigned exam grades using a flawed model after the government scrapped the sitting of exams during the coronavirus pandemic. (It was later forced into a U-turn on the issue — meaning students got their (human) teachers’ predicted grades instead.)

Given so much AI-fuelled ugliness and algorithmic mistrust, you’d be forgiven for thinking there are no more quick wins left. But walking routes app Go Jauntly may have found a redeeming use-case for AI to lift app users’ spirits in 2020.

It’s beta launched an algorithmically powered routing feature that recommends “green routes” within the user’s vicinity — meaning the leafiest and most pleasant/scenic (i.e. less polluted) urban walks possible — drawing on its understanding of users’ walking behaviour. The thinking being that COVID-19 lockdown-hit Brits could do with some nice new spots to stretch their legs locally and enjoy a change of air.

Go Jauntly’s app has been around since 2017, with more than 175,000 downloads of the (free) app to date, but it’s hoping the algorithmically powered green routes will be a game-changer for scale — given all walks in the app have been manually created by actual (human) boots on the ground up to now (including some user-submitted walks).

That said, the feature is only available to users of the app in the UK and Ireland (and only on iOS; Android is due to get it next Spring) — but the plan is to roll it out globally later in 2021. (The rest of Go Jauntly’s app is currently also available in Sweden, the US, Canada, New Zealand and Australia.)

As well as recommending the most scenic/least polluted route to walk between two destinations in the UK and in Ireland, the algorithm can suggest routes that start and end at a single location — for walks lasting from 10 minutes up to 2+ hours in length.

The machine learning tech powering the green routes feature is drawing on external sources of environmental data including the Tranquil City Index (which maps London based on measures associated with tranquility, e.g. lower pollution and noise), as well as OpenStreetMap and GraphHopper data for routing.

Go Jauntly is keen for beta testers to pull on their hiking boots and road test the algorithmically programmed walks to feed in data to help its models improve over time. So it’s quite possible that an AI’s (data-bounded) notion of ‘scenic’ may not live up to your human standards.

Trusting an AI’s urban walking route recommendation could also mean you end up passing through a less nice and/or welcoming neighbourhood than you’d expected.

Or you might find your route barred because the app is erroneously suggesting you walk through private property — much like a satnav trying to send a car the wrong way down a one-way street.

Ergo, green route guinea pigs should keep their eyes peeled — and definitely avoid straying into pastures new that contain cattle.

Go Jauntly says it hopes to continue to develop the algorithmic feature to incorporate more data sets in the future — such as accessibility information, toilets, and historical points of interest — to expand the types of route requirements it can support, working towards what it dubs a “full cross-platform digital ‘nature prescription’ in 2021”.  

It monetizes its hike-loving community of users via an optional premium subscription which gives access to extra content such as curated walking routes and guided tours, as well as the ability to download certain types of content such as walking trails for offline use.

Responses are currently closed, but you can trackback from your own site.

Comments are closed.