Emily Blythe: Pyper Vision Founder on the new fog prediction model for the airline industry

The days of having your flight delayed by fog could soon be over. 

Every year, the airline industry loses a collective $13 billion as flights are delayed or cancelled due to foggy conditions.  

Pyper Vision is aiming to solve this problem, with new technology that predicts fog and prevents delays. 

Founder Emily Blythe told Mike Hosking that they’ve built an initial model that can predict fog more accurately than traditional models, although they it still has a way to go in terms of lifting performance. 

She says they’ve teamed up with a team at Attentive Technologies to collect the data sets they need to improve accuracy. 

Blythe says it should be making an impact for the public in early 2026. 

LISTEN ABOVE 

See omnystudio.com/listener for privacy information.

Episode Transcript

Speaker 1 (00:00):
So a fun fact for you, each year of the
airline industry globally loses thirteen billion dollars to fog. So
we now have technology that predicts fog, predicts fog and
prevents flight delays. Local tech piper Vision are. The founder
was Emily blythe who's with us?

Speaker 2 (00:13):
Emily, good morning, Good morning, Mike cal Y.

Speaker 1 (00:16):
I'm well, thank you. So this is in the sense
that you're predicting, not trying to clear, right, So there's
two ways of going about this, that's correct.

Speaker 2 (00:24):
Yeah, predicting is going to be our first stage an approach.

Speaker 1 (00:27):
Now it's where are you at in the process. I mean,
if you've got a product set to go, you close
to where are you at?

Speaker 2 (00:32):
Yeah, So we've built an initial forecasting model that can
predict fog much more accurately than what traditional models can.
It's still got a lot to go in terms of
lifting performance. We need additional data sets to plug inn there,
and we've turned up with the team ad Attentive Technologies
to colleck that data, which is really exciting. But we
should be ready to have or start making impact for

(00:55):
the public early twenty twenty.

Speaker 1 (00:57):
Six with what level of accuracy.

Speaker 2 (01:01):
So at the moment, with existing data only, we're performing
at a fifty percent jump on traditional models, So New
Zealand accuracy is around nineteen percent. Traditionally we're at thirty
two percent with existing data only, but we're going to
be adding in a lot more spatial awareness at ground
level two which will help lift that performance much better.

(01:21):
So I'm hoping to get way better than a queen flip.

Speaker 1 (01:25):
So nineteen percent currently, In other words, you've got a
far greater chance of being completely wrong.

Speaker 2 (01:32):
Yeah yeah, gosh, So on those days where fogs yeah
due to form?

Speaker 1 (01:36):
Yeah yeah yeah. Is generative AI involved that? Does that
solve everything? Or not?

Speaker 2 (01:42):
Not quite. What we're using is we're wanting to see
Traditional models are using numerical weather models to predict fogs,
so that relies on our own scientific understanding of how
fog forms at a local level, and when our own
understanding is not very great, the calculations and mess that
what the model can achieve isn't isn't very good either.

(02:04):
So by switching through to a machine learning model, we
can actually start to kind of harness AI to teach
us what are the patterns that we're recognizing, and then
we can intupulate that data to actually learn more about
sold longer term.

Speaker 1 (02:18):
How in advance can you potentially do it? Because it's
the in advance but that I'm assuming allows an airport
or an airline to react.

Speaker 2 (02:27):
Yeah, so the airline schedules are set up to respond
super super quickly, which is great. And we're aiming initially
for a three hour forecast horizon with real time data,
so providing sort of minutely updates coming through just to
give a bit of context there. Current forecasts are only
updated every three to six hours and so that all

(02:50):
gives them a lot better information around whether or not
those flights can change, and within the New Zealand market
it should. It will give the airlines in our time
in order to make those flight changes needed, and we'll
slightly grow that up from there.

Speaker 1 (03:04):
Super exciting. I'm going to get you back in twenty
twenty six and we'll have a good talk about it
when it's all set to go.

Speaker 2 (03:08):
How's that sounds great? Thank you?

Speaker 1 (03:10):
Hardly to talk to Emily go well, Emily blythe I
don't know how old she is. I do know how
old you. She's in a twenty she's another one of
these young new Zealand freak shows your parents would be
unbelievably proud of. But talk about rip the top off
that we've all always suspected, which is what they haven't
got a clue. They haven't the slightest idea. Next time

(03:31):
you see somebody predicting exactly there might be fog patches
tomorrow morning, nineteen percent, just turn off your TV exactly right.
You think about the amount of broadcasting time that's been
given to what we now know to be literally nothing
more than guesswork, and nineteen percent charts a big It's
about as accurate as the profit made by the Prime
Minister on a house. Welcome to Media, twenty twenty four.

(03:54):
Styles For more from the mi Casking Breakfast, listen live
to news talks it'd be from six am weekdays, or
follow the podcast on iHeartRadio