and this week, we put together the best of the stories we covered there. we're bringing power to the flower with the drones and ai dealing with the changing climate of this blossoming industry. we're also in the rural communities where ai is improving prenatal care. and we explore the piles of electronic waste making locals choose between lives and livelihoods. agriculture plays a huge part in many people's livelihoods in kenya. in fact, the industry employs more than 40% of the total population. but climate change is having a real impact on the crops there. and you might be surprised to find out what one of the most sensitive crop types is. flowers. they have a specific window for being grown and sold, and if the weather's out of sorts, it can completely derail the process.

but technology can help. and shiona mccallum has spent a day on a rose farm in naivasha to find out more. our years are punctuated by times that we want to buy flowers — valentine's, birthdays, christmases. and the chances are, when you look in your bouquet, one of the stems would have been grown here in kenya. but like most places, this part of the world is not immune to the problems that climate change is bringing. rising temperatures, extreme droughts and flash floods are affecting the crop yields. in fact, just a couple of years ago, some kenyan flowerfarms reported losses as high as 50% due to the impact of heavy downpours. so that's where the technology comes in. artificial intelligence company lima labs is providing a machine vision system using drones, which they say makes the farming smarter

and more predictable. we use a camera on the drone to take images of crops in the farm. and then once we've gotten those images, we send them to the office and then we're able to process them and give insights to farmers. and insights span all the way from a prediction of their harvest, it spans all the way to a prediction of stem density, growth rates, probably in the future, even detection of pest and diseases. on this farm, some of the flowers are grown inside huge greenhouses. the rest are outside — and that presents some challenges for the drones. with more gusts of wind, the drone uses actually more battery power and drains a lot of battery power. and also the rain — we cannot fly the drones through the rain. the ai algorithms being used can help predict the weather patterns through sensors

on the farm that detect humidity, temperature and other environmental conditions. a dashboard on a computer screen displays footage of the plants, providing stem and flower head count and other crucial data, like the chance of an insect infestation. this all helps farmers better predict fluctuations in growing and harvesting patterns. well, the big issue is climate change and the unpredictable weather patterns. the heavens have opened now, so we're going to head inside, into the packing room, and do the rest of the interviews. hi, i'm shiona. hi, i'm mercy. nice to meet you. nice to meet you, too. can you show me around? sure. 0k. tell me what's happening here. this is our grading hall. it's one of the - post—harvest sections. post—harvest, that means. we process the final product that is from harvesting - to packing it and dispatching to the final customer.

and can you talk to me about some of the varieties? because these lovely little red berries are nice. these specifically is - hypericum, the red berry. so the beauty of this crop lies with the berries, - the red, shiny berries. the quality of these berries and the leaves looks fantastic. would you say the technology is really helping them be in this condition? yes, it is really helping - because with the technology, we have been able to come l a long way in producing more strong varieties, - more resistant ones. how has the technology been received here on the farm? what does everyone think about having drones in place and using the tech? we didn't believe in it at first, but later on, i when we saw the results i and benefits, it's now a way of life on our side. so from there, the farmers conduct weekly farm assessments. depending on the flowers. health, they can change their sales and logistics decisions, meaning they don't

waste time or lose money. another thing - is time management. for the shortest time| possible, and you get more accurate results. i can imagine people will be pleased by that. yeah, yeah. could you talk to me through how a worker's day might have changed during the nine—to—five, as a result of having this technology implemented? so what they used to do is get a square metre and manually count how many stems are there in that square metre and then extrapolate and assume that the rest of the field is actually similar to that square metre, which obviously is very inaccurate. with lima, when we come and we monitor these crops using the drones, we are able to, first of all, get accurate numbers because we are seeing each and every stem and we don't have the human aspect of getting tired, right, to count or losing track of numbers, right? there's that aspect. and then at the end of the day, we're able to get an accurate number for the amount of stems in the whole field. this, of course, saves the four hours the person was doing

that every day. at the end of the day, makes their life a lot easier. and back at the office, they can see the results. because the reality is here, | it's showing you the number of stems that you forecasted and the actual stems - because this is the reality. yeah. so very confident with| this technology, yeah. technology is rapidly changing agriculture in many ways. and this farm is certainly taking steps to move beyond the traditional and into a data—driven business. a woman dies every two minutes due to complications in pregnancy or childbirth. the majority of all maternal deaths are in sub—saharan africa. and in kenya, the problem is actually getting worse. 70% of the population lives in remote rural areas, where they're cut off from

life—saving infrastructure. not all mums are able to access the right kind of care. transport, getting to the hospital is always an issue. we don't have enough health care workers for the population. big problems, i would say. not all maternal health clinics across kenya are as well—equipped or as well—staffed as this one in nairobi. but it's hoped the technology being trialled here could help ensure that more mothers—to—be get access to potentially life—saving treatment. this is janefer�*s first glimpse of her baby. the legs. oh, wow! ultrasound scans like these play a key role in ensuring a pregnancy goes smoothly. unfortunately, there's been some bumps in the road for janefer.

last month, i had a complication. i had a discharge. she was advised to go straight to hospital to get checked out. i was scared. yeah, i was scared. i was like, "maybe i'm losing the baby or something." luckily, janefer and her baby are both fine, but she also had problems in her previous pregnancy. i had to go for caesarean section. and did your doctors tell you that it could have been prevented if you'd had an ultrasound earlier? yeah, they said it could have been prevented. where you live, is it difficult to get an ultrasound during your pregnancy? it's quite expensive. and you have to go for a distance to get it. jacaranda health, the organisation that runs this clinic, has partnered with tech giant google to trial a solution that could help save lives. it's important for pregnant women to do ultrasound throughout their pregnancy.

the first importance is to check foetal anomalies. initially, as a nurse, we were not able to perform the ultrasound. until now. this ultrasound probe sends video output to a tablet or smartphone, making this system cheaper and much more portable than traditional ultrasound scanners. artificial intelligence then interprets the image without the need for a sonographer. the information that i'll get here will help me to save the life of a mother and also save the life of the foetus. i've come to google�*s offices in nairobi to find out more about how their ai models are being developed and how people who aren't sonographers can be trained to use it. if this technology is intended to be so easy that you could train anybody to do it, do you think that i could have a go? i think you definitely could.

first, i need to measure the size of the bump and apply my ultrasound gel. so as you smooth it, you'll start to see the imagery. oh, wow. yeah. next, it's time to do the scan. i need to do six sweeps over the abdomen, and the app gives me instructions for every step of the process. you're doing amazing. thank you. feels like if i were more confident, this would be much quicker. the ai models then analyse the image. they give two vital pieces of information, the gestational age — that's how many weeks old the foetus is — and the foetal position, or the direction the foetus is facing inside the uterus. so it says gestational age is 23 weeks and two days. the foetal presentation is... i actually can't pronounce that. cephalic. what does that mean? thatjust means heads down. non—cephalic positioning could be breech or transverse,

which means that the patient might need a c—section or something like that. we're not trying to replace sonographers. humans are very important to providing care, and we just want to give them additional tools in their tool set. around 100 scans have been collected in clinical settings in kenya so far. a lot more than that will be needed before the system's ready for routine use. although the research is still in its early stages, jacaranda health hopes the trial marks the beginning of better days for maternal health in kenya. 10% of mums in rural kenya have access to ultrasound services. if this is available, it will definitely be a game—changer.

time for edtech newzeal. it mimics the human artist like this, voices, and sound. artist unions have sought to pressure the regulation of the use of ai. studios had become interested in reducing production costs. meanwhile, the uk and us have signed a landmark deal to work together on testing advanced artificial intelligence. both countries have agreed to work together on developing robust methods for evaluating the safety of ai tools and the systems that underpin them. nhs england is to offer tens of thousands of people with type i diabetes a new technology dubbed an artificial pancreas. the device continually monitors a person's blood glucose levels and automatically calculates and adjusts the amount of insulin given to them through a pump. and the world's most powerful

laser is now operational at a european research facility in romania. the system achieves ten watts of power, that's a lot of power, in under a femtosecond. that is i,000,000,000th of 1,000,000th i,000,000,000th of 1,000,000th or i,000,000,000th of 1,000,000th or one quadrillions of a second. just keep away from me! now, there are around 25 diseases for which we have safe and effective vaccines. but millions still die every year because those vaccines don't get to the people who need them. so one company has been looking at ways to streamline the process and, in doing so, save lives. chikungunya is one of the infections that's spread by mosquitoes. it's usually confined to sub—saharan africa, but has started to spread.

one vaccine has just been approved by the food and drug administration in the us, but others are working on it too. one of them is the bristol—based start—up imophoron. it's developed a new synthetic protein called addomer. the protein is adorned with parts of the target virus, but it's not infectious itself. it's a different approach to how other vaccines, like the covid—i9 ones, were made. but there's one big advantage. the addomer can be produced, stored, transported and brought to the people at ambient temperatures, which can go up to 50 degrees. this is a big advantage to other vaccines, which depend on a functioning cold chain. millions of doses of vaccines are lost every year because of inadequate refrigeration, especially when being sent to developing countries. part of the issue is- transporting the vaccine from the distribution depot, | where it is kept in cold chain conditions, to the site where i it's going to be administered.

it could be hundreds of miles. and then when they get - to the facilities where they're going to deploy the vaccine, more often than not, - there's not any refrigeration. now, over the years, - people have tried different ways of dealing with this. they, for example, would dig holes and put the vaccine - underground to try and keep it at a sort of temperature. - but it is very, very hard. another big advantage of the technique being used here is that it could be synthesised to target more than one disease in one shot. the addomer is not yet in clinical trials, but it couldn't have got to this stage without two key technologies. we are now coming to the cryo—electron microscopy facility. this is where our structure data collection will take place. this is ufuk. he is preparing the experiment. hello, i'm lara. nice to meet you. a prepared sample is placed in a tiny drop of water and inserted into that electron microscope at minus 195 degrees celsius.

this so when we put it into the cryo—electron microscope, the electron beam can go through it and we will see our vaccine particles. 0k. can i take a look at what it's actually seeing now? so we can only have one of those grids loaded. . those are the three millimetre disks that we applied - the sample onto. this is a magnified image of that. - so this is a grid square image. and now we're going to zoom into these foil holes. - the microscope can zoom into 0.1 nanometre detail. that's one ten—millionth of a millimetre. these raw images may not look like much to you or me, but they are a vital part of the process. and ai is pivotal to this. where ai can help us is, we can run literally thousands of simulations of where to position which particular bit of the protein to attach, where to put it, or if we need to stabilise it or fix it in some way. and that can help us really reduce the time in the lab

spent doing iterations until we get the right molecule. but if they do succeed, the possibilities in other areas could be huge too, such as cancer vaccines that train the body's immune system to make antibodies against a tumour. and at least its methods do have efficiency on their side. the addomer uses established technology for its production. i the manufacturing process - are established and we believe that they will be i very competitive. and we are now working towards entering clinical trials. _ this is a small company, though, up against a big challenge and big pharma. and these vaccines aren't in humans yet, so there's a way to go. have you ever wondered where your old laptops, phones and tvs go to die? well, every year, our planet generates over 50 million

tonnes of electronic waste. but only a quarter of that e—waste is properly recycled, which is really bad because our devices are full of potentially toxic chemicals. michael kaloki reports on how one company in kenya is hoping to combat the issue. this is dandora. this community on the outskirts of nairobi is home to one of the largest dump sites not only in kenya but east africa. although it was declared full 20 years ago, over 800 tonnes of garbage is still dumped on this sprawling 30—acre site every day. and in recent years, an increasing proportion of this is e—waste. men, women and even children scour the stacks of waste for discarded electronics, breaking them apart for their motherboards, batteries, wiring, anything that can be resold for parts.

it's become an important income stream in what is a desperately deprived area. i've come to this part of the dump site to meet an e—waste collector. there are thousands of people who work in the dump site, and it's a dangerous environment to work in. stanley mwangi is one of them. translation: we get a lot of e-waste here, | as you can see. we find items like motherboards. we repair some of the motherboards, and others, we just use as replacement parts. anything we can't use, we sell to chinese traders. there are many buyers for us materials. but e—waste is not easily found. the items are few and the buyers are many, so the price shoots up. we sell the e—waste by the kilo. traders pay 1,500 shillings,

which is about $6, per kilo. he's worked on the dump since 2002 and, as a single parent, says he has no other choice. translation: it is dangerous work, but we have to do it. . if we do not, we will go hungry. one of these fingers got chopped off working here, but i had to keep working here to make ends meet. he says the injury was caused by a shard of glass. and while such injuries are common, where people don't have access to any protective gear, there are arguably greater dangers to working here. they are exposed to toxic chemicals such as lead, mercury and cadmium. these chemicals can leach into local water supplies and noxious fumes inhaled when

any unwanted e—waste is burned. while the informal and partial recycling of electronic waste has been the status quo in kenya for decades, one organisation is hoping to change that. so, some of the waste from landfills, like the one in dandora, ends up in e—waste centres such as this one. well, i'm here to find out what the process is when the waste gets here. the waste electrical and electronics equipment centre collects and recycles a variety of e—waste, ranging from computers and phones to solar panels and refrigerators. at its nairobi headquarters, these are then channelled to the relevant workshop. in one, technicians test and refurbish laptops and tablets so they can be used again. and any device or appliance deemed beyond repair is sent to their dismantling room to be stripped for usable parts. and in another room, hundreds of batteries

are upcycled, saving them from contaminating the soil in dumpsites like dandora. at the moment, we are able to receive 240 tonnes of e—waste per year, though we have the capacity to process 200 tonnes per month, which means we are performing way, way, way below capacity. so why, then, is there this low volume of e—waste being collected? it's due to lack of public awareness. majority of our community members do not know that e—waste is dangerous. so, joseph, is all the e—waste that is found in kenya, is that generated here? the e—waste that is produced is as a result of electronic devices that have been imported into this country. but of course, we also have electronic devices coming to this country disguised as donations. you can't use them for more than even two months. but despite these hurdles and

the prediction that the amount of global e—waste could double by 2050, joseph hopes that organisations such as his could start to make a positive difference. but what difference that could have on the livelihoods of people like stanley who rely on dumps like dandora to survive is now uncertain. translation: i've been able to open a repair shop - with the money i've made here. so i will keep going. i will only stop working i when my body gives up. that is all we have time for. yeah, thanks for watching and we will see you soon. bye!

hello. storm kathleen will dominate our weather through this weekend. a deep area of low pressure passing well to the west of the uk, but close enough to bring gales in northern and western parts. one other consequence though — some very warm air being drawn up from the south. here's how the storm developed during friday. a strong and powerfuljet stream really energising this area of low pressure, this curl of clouds developing here. wherever you see a weather system spinning up into a curl like this, well, it shows that it really will mean business. this deep, low passing to the west of ireland, lots of isobars squeezing together. so some very strong winds, but also some very warm air. northern scotland, far warmer than it has been over recent days. parts of eastern england could get to 21 or 22 degrees. some rain to start off across the northern half of scotland — heavy rain at that. it will pull away northwards and then essentially

it is a sunshine and showers day. some of those showers could be on the heavy side, but it is going to be windy for all, particularly around these western coast. gusts of 50, 60, 70 miles per hour in exposed spots. so that could cause some disruption, certainly some very rough seas as well. but with those winds coming from the south, yes, it is going to be warm with temperatures peaking at 21 or maybe 22 degrees in eastern england. now through saturday evening, this weak cold front will exit the east of england. so that will leave some slightly fresher air into the early hours of sunday morning. still windy, still with quite a few showers. not as mild to start sunday morning, but still very mild for an april morning. and then into sunday, well, it's another sunshine and showers day. the showers likely to join together into bands. so some places could see quite a few of those heavy and possibly thundery showers. windy again, strongest winds,

this time in the north west of scotland. gales or severe gales are likely still warm, but not as warm. temperatures north to south between ten and 16 or 17 degrees. but then as we head into the start of the new week, we're watching this area of low pressure. now, there's still a bit of uncertainty about this, but this could bring another bout of wet and very windy weather to southern parts of the uk as we head through monday and into tuesday. we'll keep you posted on that one. whichever way you slice it, an unsettled week of weather lies ahead. that's all for me. bye for now.

live from washington. this is bbc news. two senior israeli military officers have been fired after seven aid workers were killed in missile strikes in gaza. an earthquake rattles new york and parts of the northeastern us. plus: president biden tours the the site of a collapsed bridge in baltimore, as a third body of one of the victims is recovered. all the families and loved ones

were grieving. i've come here to grieve with you, we all are. hello, i'm carl nasman. the israeli military fired two senior officers on friday — after reviewing monday's airstrike on an aid convoy in gaza that killed seven aid workers. the idf said the airstrikes came after a series of "grave mistakes." it also gave an account of how and why it carried out the attack. it said in part that the strike should not have happened and expressed its deep sorrow for the incident. it says those who approved the strike were convinced they were targeting hamas operatives. the strikes on the three vehicles were carried out in serious violation of standard operation procedure and that the brigade chief of staff, an officer with the rank of colonel in reserve, and the brigade fire support commander with the rank of major have been dismissed from their positions. our middle east correspondent lucy williamson has more on the fallout from ashdod, israel.

israel says hamas gunmen often pose as aid workers.