Deep data dive: how busy are the BA lounges at Heathrow?

Analysing live and historic lounge occupancy rates at LHR

BA has a public-facing webpage that displays occupancy levels across four of their lounges. By recording this data over a period of time, and unearthing hidden data on a fifth lounge, I’ve been able to do some interesting things with it, including an expanded view on live occupancy levels, and some analsysis on historic data.

How I’ve created this data

Firstly I”ve analysed how busy the lounges are by hour of the day, and day of the week, to visualise the hotspots for each lounge.

I can also use that data in real-time to show you not only how busy the lounges are (with an actual headcount, versus the BA site, which only shows a percentage of how full they are) but also give some commentary on whether it’s busier or quieter than usual, and what’s expected to happen in the next hour.

Finally, using this data, you can enter a day of the week and time and I’ll show you how busy the lounge is expected to be, and if it’s worth arriving a little later or earlier.

How busy are the lounges right now?

Predicting how busy the lounges are for your next trip

Enter your expected arrival time and day of the week, and I’ll show you how busy I’d expect the lounge to be at that time, plus a view on the occupancy just before and after your expected entry time.

Analysing the LHR lounge usage

Occupancy by day of week and time of day

How does BA track occupancy at their lounges?

British Airways appears to use a company called LiveReach AI, who offer “video intelligence” services. That suggests that there are cameras in the lounges that track you as you enter, and then as you leave.

Looking at their site, it’s a clever solution that helps airports understand what’s humans are doing in their buildings. They also cater to the car wash industry, schools and regular retailers.

They’re also claiming that their service does not require facial recognition, which suggests that BA are not storing details about who you are.

There are issues with the data

Recently there has been abnormally high occupancy rates in the early hours of the morning in one lounge. At first I thought this could be staff members, perhaps a late night refurb or deep clean, but the numbers are too high (on one day the lounge had 171 people in it at 1:50am).

There is clearly an issue with the sensor or the software in this lounge. The 171 reset to 0 at 2am, which would suggest that there is a hard reset at that time every day.

To cleanse the data I’ve used a query that detects days where the occupancy is abnormally high, and those days are not included in the heatmap and charts above.