let me take you up. up into the air. up above the clouds, above the weather, above all the aircraft. this is the stratosphere — a place yet to be conquered by humankind. up here, the air is thin and calm. and it is here that you'll find the zephyr. this is a strange beast — and the fact that it flies this high is the least strange thing about it. see, it only travels at aomph, it only weighs 75kg.

it's launched by hand, it's completely solar—powered — and in theory, it may be able to stay up here for months. i was last at zephyr�*s base in farnborough in the uk in 2018, when it had just stayed aloft for very nearly 26 days. since then, it's done 64. so, this is how it works. during the day, the sun hits the solar panels, which charge the batteries and power the propellers, and the plane climbs to 75,000 feet. when the sun goes down, the batteries completely take over. the propellers do slow down and the plane does lose altitude. the trick, though, is to make sure you're still above 60,000 feet by the time the sun comes up the next morning — and the process can start again. so, will the entire wing be covered with solar cells,

every single square centimetre, as much as you can in solar? actually, the solar array that we use now typically is so efficient that we don't tend to need to do complete coverage. really? yeah, we can actually recharge the batteries most normal days by lunchtime. wow! in theory, is it possible to stay up forever in this? well, i think eventually, we'll get as close as we can to that. at the moment, our limitation is the number of cycles the batteries can cope with. so, a cycle is a day — so a full charge, and then a full discharge, that's one cycle. and we're targeting six months in the stratosphere at a time, and that's in line with the battery performance that we see coming through. so, look, i get it. i get that this plane can stay up in the air almost indefinitely. you only have to look at it to understand that it's something pretty special. the main question i have is why?

from the stratosphere, at 60—70,000 feet, we can talk directly to a standard mobile phone, so the aircraft will function exactly like the cell tower that you have today, but it's high up. and, because it's so much higher up than the regular cell tower, it can cover the equivalent of about 200 base stations on the ground. so, of course, that replaces notjust the, you know, the equipment on the tower, but it replaces the whole tower. an uplink antenna would connect the mobile network to zephyr, and then zephyr could serve an area of 7,500 square kilometres, in theory, for months at a time. this is a real proposition. or in the event of, say, a natural disaster, a plane could be flown to an area that had suddenly become cut—off from the grid. mobile operators have something called cell on wheels — cow — which is usually a truck that has some equipment. we have a cell on wings, a flying

cow, which is flying cell on wings. these cows — er, planes — will fly autonomously. there are no joysticks involved. the pilots send the planes coordinates and flight plans — but they're mainly interested in how much energy it's using, and making adjustments to keep its batteries fully charged. so this is the ground, and this is 80,000 feet. and all this colourful stuff, that's wind — you don't want to be in that — but if you can keep the zephyr above it, from about 60,000 feet and up, it's pretty plain sailing. the problem is, you've got to get to 60,000 feet in the first place — something that can take ten hours from that weird hand—launched ta ke—off. we've spent a lot of time and effort studying meteorology in the tropopause, so that we can understand how to transit through it as safely as possible with this aircraft. and we've now done a global study of where all the weather is, and all the different weather

conditions around the world, and we've found some of the best locations in the world to start launching and landing these from regularly. so, where the air is nice and clear, all the way up. all the way through, yeah, and once we're in the stratosphere, we're away. in fact, zephyr�*s parent company has just announced kenya as the location of its first planned permanent launch site. but we have seen these kind of ventures before. for example, google�*s loon project was also based in kenya, and its balloon—based mobile cell tower project failed to stay afloat. and what about satellites? these days, very small cubesats can be put into orbit at much lower cost than their bigger siblings — and they stay up automatically, no power involved. although, unlike zephyr, you can't bring them down and swap out their payloads. with satellite, if you want to talk to a handset, it's very difficult. and if you manage with low earth orbit satellites,

you can talk to a handset, but it would be very limited. you know, it's sms, maybe a few kilobits per second, you cannot do full 56. you're not mimicking what a terrestrial station does. and if you have satellites that are really big enough that can do something like this, the economics are so expensive — and then, you're spreading them across the whole planet, you're not getting the efficiencies, you cannot scale. so that's on the connectivity side. 0n the earth observation side, satellites are great because they can view anywhere on the planet, but not persistently. because they take an image, come back after a period of time. with high—altitude platform stations, you can do that persistently, so you can see the change over time. and there is competition in the stratosphere itself. similar aircraft are being developed by several companies — including this one by bae systems. whatever the final use for these so—called high altitude platform stations — haps — there does seem to be both the appetite and now the technology to fly high and stay high.

birdsong paul: this is the buckthorne cutting nature reserve, home to dozens of species of plants, insects, and animals. it's hard to believe it's just a few kilometres from the very centre of london, and a technology hot spot. it's a haven of calm and serenity... ..until this happens. love a train. the area here is based on previously unmanaged network rail land,

the body responsible for most of the sprawling railway network in great britain — over 52,000 hectares. they're using a mixture of traditional technology alongside machine learning to identify and monitor wildlife using the space here. we've got a massive estate across the whole of britain. it's 52,000 hectares which, if you squash it all together, is the isle of wight—and—a—half. and that estate, that network goes through every sort of habitat that you can possibly have. and so, we've got that biodiversity on our estate, and we can manage it and work with it whilst we're running trains. so i guess, with the kind of information that you're collecting, the research that you're doing, how is that beneficial to network rail as a company? what kind of things can you do with that information that you find out? it helps us monitor the condition of the habitats that we've got, so we can use that when we're trying to achieve our biodiversity targets. but it also lets us see what species are using those habitats. and if we know what's there before we come in and do any work, then

we can start mitigating for that. we can start doing the work at the right time of year or at the right time of day, so that we're less of a disturbance on those species. the monitoring is being carried out by zoological society of london — or zsl. they're using motion sensing camera traps to help identify some of the animals. this is a camera trap, and we use this for monitoring wildlife. typically things that move, that are kind of larger—bodied, and that move along the ground. so we'll typically put it kind of on a tree, around this height — and there's a camera here and there's a sensor. so, if a warm—bodied animal moves past, it will trigger a number of photos. oh, look, that's us — hello. in the uk, we're using it largely for ground—dwelling mammals — so like, foxes and badgers and squirrels. as well as camera traps, the monitoring team also use specially designed acoustic monitors to capture audio from birds, bats, and rodents.

while the method of collecting the data may be relatively low—tech, what happens with it afterwards is very much cutting edge. zsl�*s initial pilot project captured 35,000 data files — or 3,000 hours of audio — from just 33 acoustic monitors placed across network rail's estate in south london alone. they needed a way to efficiently analyse such large volumes of data — which is where our old friend ai comes in. zsl worked with google, using cloud storage and pre—trained machine—learning models, to identify and map species to the network rail estate. back at zsl hq, i sat down with lydia to see some of the results of the analysis. in some cases, there are some really good machine—learning models already trained — so for instance, for birds and bats, there are some excellent models out there. and essentially, they use these

spectrogram images that i'm showing you here — these are really distinctive images — they basically learn these images which are associated with these different species. the ultimate aim for this research is to understand the true scale of the biodiversity around network rail's 20,000km of railway corridor. and advances in technology like this make that more achievable. spencer: and that's it for the short cut of click this week — the full—length version is waiting for you right now on iplayer. thanks for watching, we'll be back next week. see ya.

live from washington, this is bbc news. israel's military says it recovered the bodies of three hostages abducted from the nova music festival on october 7. the us military confirms the first deliveries of aid have arrived into gaza via a new temporary floating pier.

with heavy fighting in ukraine's north—east, we have a special report from the front line. ukraine is just about holding the front here in the kharkiv region while the russians have launched this fresh assault. hello, i'm caitriona perry. you are very welcome. israel says it has recovered the bodies of three hostages from gaza. the israeli defense forces said they died during the october 7 hamas attacks where around 1,200 people were killed. the victims are itzik gelernter, shani louk and amit buskila. until now, it was hoped buskila and gelernter were still alive. shani louk was confirmed dead at the end of october. the israeli army believes all three victims were killed at the nova music festival. they were among more than 250 hostages taken into gaza, of which roughly half

are still unaccounted for. the idf expressed condolences to the victims�* families.