An app for avoiding busy supermarkets

08 June 2020

Interview with 

Yohan Iddawela, London School of Economics

EMPTY_SUPERMARKET

A supermarket aisle empty of people.

Share

As lockdowns start to ease, public spaces like supermarkets are likely to become more crowded. That’s a worry if you’re at high risk from COVID-19, or you live with someone who is and you’re still trying to shield yourself. The best solution is to go shopping when the stores are quietest. But how do you know when that time is? Step forward LSE Geography PhD student Yohan Iddawela, who told Chris Smith about his app for just that…

Yohan - Crowdless is a free app that shows you how crowded supermarkets are, so you can choose a less busy time to attend or a less busy alternative.

Chris - And how did this come into being?

Yohan - It's a funny story actually. We were initially designing a product to help people navigate safely in conflict zones. We were trying to provide real time security alerts to people in conflict zones, so that's when and where a violent incident like a bombing, shooting, or kidnapping took place, so you could avoid violent hotspots and stay safe. We were trying to roll it out in Colombia, when Colombia went into complete lockdown because of coronavirus. So we repurposed the same technology to help people navigate safely during the current pandemic.

Chris - So from Columbia to COVID! How did you manage to repurpose the app to do this?

Yohan - We were using some of the same underlying technology. It's called crowdsourcing technology, where people would input information directly into the app. And that forms a core part of Crowdless.

Chris - Can I download that? Is it in the App Store?

Yohan - It's free to download on the App Store and the Play Store.

Chris - And what am I searching for?

Yohan - You're searching for Crowdless.

Chris - Okay. Crowdless. Okay, here we go. Install. It's downloading.

Yohan - The moment of truth...

Chris - It's asking me "can it access my location". What's that all about?

Yohan - That's about showing you location-specific results. So you want to see supermarkets in your vicinity, and in order to do that we need to know where your location is.

Chris - Okay. So it's started up and there's a map being displayed to me.

Yohan - So that map will show you all the supermarkets in your vicinity.

Chris - Okay. I know where there's a supermarket that I usually visit a few miles from me.

Yohan - Great.

Chris - Right, what's now coming up on the map are big green blobs. Are those the supermarkets?

Yohan - They are. Yes. And green means they're not so busy at the moment.

Chris - Right, so this has located for me a Co-op, and it says the estimated crowdedness is 27%. Now what does that actually mean?

Yohan - A hundred is as busy as it ever gets. And zero means that there's no one in there at any given time.

Chris - The other thing is asking me Yohan is it says "contribute". Does that mean that if I go there, I can now say, "well I'm in here and actually it's heaving with people, the app is wrong"?

Yohan - Exactly. Or alternatively you can confirm the rating. And the more people that use the app and contribute, the better the data is going to be.

Chris - Now how are you actually getting this data? Is it purely people like me walking into that particular store and saying how busy it is? Or are you getting other sources of data to inform this? Because to be honest, I'm quite surprised it found a co-op near me. I live in the middle of nowhere.

Yohan - Co-ops are everywhere, I guess. So we get third party data - that's from services like Google, as well as sensor analytics companies. And these are companies that install hardware in stores that track footfall. So you can think of that as a base layer of data, and then we add crowdsourced data from our users, so similar to the contribute button you were playing with. And that adds a certain degree of accuracy to the data.

Chris - I'm a bit surprised to hear that organisations have already got monitoring in stores to track footfall. How common is that?

Yohan - It's quite common actually. So it helps stores optimise where they place certain items. So to give you an example, a store might hire a sensor analytics company to track where their customers are first going. So if they see that their customers are going to the bread aisle first, they might put the bread aisle at the back of the store to encourage people to walk past other items. So all we're trying to do is surface this internal data and make it broadly accessible to the public.

Chris - And where do I stand now that I've got this running on my phone; it clearly knows where I am because it homed in on my geography; and even when I tried to rate a store just now, as you were talking, which is nowhere near me, to contribute to what I thought the busy-ness was, it said I was too far away and couldn't possibly know! So how much do you know about me now that I'm running this?

Yohan - We know almost nothing about you. So you might also notice that there's no login screen. So we don't actually know who our users are. We just use the location data to provide you with location-specific results, but we don't store that data anywhere.

References

Comments

Add a comment