 So thank you to all the speakers. I'd like to invite them all now on to the stage so we can have our group discussion So any questions? We'll try and take them in groups before presenting them to the panel So we have two questions down here and one so if we take these three Hello there Armin Sprecher MSF my questions for Benjamin and I'd like to follow up on what Claire asked earlier Yes, human beings are much better image processors than most automated systems But some of them are not too bad like neural networks if they have a good training set you had mentioned how you were gonna Think about combining your data with other systems Have you considered actually using the data that you're generating as a training set for automated systems? And what do you think the potential for that is this question? Yeah? Yeah, thank you Kieran MSF now you mentioned that Missing maps has helped for population counts, but actually this is not a question necessarily for Nell It could be for Ben or for anyone else thinking about the problems with clouds and other and Recognizing makeshift houses and so forth and the fact that in the recent years. We've heard Mapping being used for estimate, you know rapid estimation of population. Where are we now with this? Is this getting better? We now actually do feel that we're now able to make really quite more reliable estimates estimations of population counts because again We're hearing yesterday that's often a really big challenge for all of our planning of our interventions We just don't know the population. So are we improving and is mapping going to be the solution? And we have a third question here Seeing maps my question is for is yet you talked about the operational costs of the SMS system Have you got any data on any savings made possible in the overall operational costs of those health centers? With improved reporting Thank you. So I think the first question was addressed to Ben Okay, so yes in the end I also Mentioned like a project where you want to use this as a training data source and does also use neural networks The thing with neural networks is that really they need a really large trained asset That's the advantage. We have now with map swipe data that we have enough data to do so the problematic thing is that map swipe training data is quite Not as detailed as you may want to have it like you would like to have a building outline That would be the perfect training data. So The final steps would then be use map swipe use open street map data combine both and then train your neural network The second question was related to better population estimates Would you like to go today? I mean I as part of the study the participants including epidemiologists commented that they felt it improved the population estimates, but I Would have to defer whether it actually has to Expertise Yeah, so maybe what I also can say to this topic is that In the first step, it's good to have the building outlines and then we can use this to estimate the population But I think it's we cannot do it really good without having data directly coming from the field to really calibrate those models And so that's why only Using the shape of the building or the area. It's maybe not enough to have really like the Detailed population information we would like to have The third question associated with the operational cost of the SMS for the ad, right? I was I was hoping either I would have your eye contact or someone will ask me a question So I can follow up on it. I could see my colleague who wasn't satisfied my with my answer to So the the cost just to to to explain it again, so we had the cost for The equipment itself. It was approximately forty thousand dollars and then we have the monthly cost for Operations in terms of SMS's and that was less than one thousand dollars a month We have done a comparison in terms of operations To the traditional system Which is sending the papers with motorcycles or motor taxi and we have done the estimates I don't have it on top of my head right now, but it was approximately three times much more expensive with a motor taxi to send these papers every week depending on the distance and Excuse me, and if it's a raining season or not as well So there was a variation and whether the paper will arrive or not also if the roads That the motor can actually drive through these roads or not the security situation So we have done a comparison, but I don't have like the data on the top of my head right now So yes, we have done that Can I take chance to have while I'm having the bottom of the microphone right now? Okay for the malaria question I made a follow-up with my colleagues With Maya regarding the malaria in fact right now Yes, I can see it. Thanks There is there is a trial right now regarding malaria for the malaria season But as I mentioned at the beginning we have used the M for this pilot that we have presented the data that we have used the requirement of MOH For the diseases, but MSF is doing something right now regarding the malaria to be tested using this system as well Are you satisfied? Thank you You cannot upset an MSF person in the room Are there any further questions? It's taking the group to have the lady there This gentleman here and a lady a bit further back So if you take those three I will start sorry. Yes, stop this lady. So this is a question for Ziyad Sorry, can you hold the microphone? Sure. Can you hear me? Okay, it's a question for Ziyad. I'm Janet with MSF France I was involved in coordinating a similar mobile surveillance project in DRC with MSF with another NGO and there we had similar Successful results in the pilot period in terms of completeness and accuracy But then over a period of about two years. We considered the project a failure partially because there was a lot of population displacement displacement of the agents who were reporting and Violence and actually a lot of the agents who were reporting lost their phones were There were problems with signal Given that you're implementing it car and it's a place that also suffers from displacement and conflict and sometimes unpredictable Unpredictably Have you guys planned for that and what were some of if so, what were some of the solutions you came up with to overcome the possible challenge of people getting displaced in the system falling down It's a very good question. Sorry, can we wait till we I'm sorry, this question here. Hi, my name is Kevin Tang I'm a master's student at the London School of Hygiene and Tropical Medicine studying public health and geographic information systems my questions for Isidoro is My questions about the future of mapping So as of right now, it seems like the data is being collected and then being analyzed or platforms are being built Outside of the countries that are being used in and then being implemented within the country themselves Do you think in the future there is the capacity for countries to be able to collect their own data through the Ministry of Health? Process that within the Ministry of Health and then implement it on the national level As these people know the country in better context and they're I mean their locals. They know it better Thank you, and then we had a third question I just wanted to know because a lot of the projects seem very Complementary and I would like to know if is this are you finding out that all of you had these projects here or Is there is there any? Has there been any back information before so that there can be a more of a Collaborative approach at different stages of the projects Thank you, and so we go back to the first question of the ad We're looking out Yes, if you looked at Future proofing the system Okay If I understood you correctly is about sustainability and especially if the system would be tested in a time of crisis In retrospect we saw that when the staff are busy for example when there is a vaccination campaign The paper reporting went really down and the SMS went a little bit down But then they can actually submit their reports over the weekend or at night because it is just have to send it by SMS In terms of Crisis I think if there is no cell network, this will be a challenge before we think about the staff themselves what happened to them in terms of costs You need to refill these Cell phones with the credit and we have used the fleet But also we have discussed that to have a toll free So even if they don't have a credit on their cell phone, they can still send to a cell phone number But it hasn't been tested but has been discussed should there be a handover to the MOH Then there is no need to manage the finance of refilling the credit or the fleet as well So they can send an SMS to a number that is paid bill will be paid by MOH So on the top of my head, I can answer this way But in terms of discussion for the future in terms of crisis there hasn't been a discussion about it But I think if there would be a crisis that that's where the system would be tested as well Thank you The next question was about whether there was the capacity within the Ministry of Health in Sierra Leone to push this project forward Thanks for the question Yeah, actually the these whole approach one of the main pillars of the whole approach was to to build capacity and to make sure that That this is something sustainable and that they can continue in the future The mapping was done with them that the validation from the field was done from from MOH The dashboard was developed always in continuous collaboration with them listening to what they would like to have We are trying to to train them So they can be completely independent in the in the use of of the dashboard our our also approach with the AP assistant The main point is like we can't leave capacity there And we are not only training him but also all the other district surveillance officers in in that district So yeah, indeed one of the main Pillars of the intervention is building capacity and trying to ensure sustainability in the long term And the final questions about whether people talk to each other So who was aware of the others projects prior to this knowing about this session Yeah, well, um, we knew I've I've heard and I've been told about these approach with the with the SMS system The thing is like again the MOH was in collaboration with other organizations at the national level and they deployed this This other system I was I was talking about but indeed we've been trying to explore other possibilities And maybe not enough and maybe there should be more communication I acknowledge that and that's what we are here Any further questions? We have one at the back and one right on the back row. We'll just take those two Yeah, meet Phillips MSF. So my question is about the use of the data for the decision-making So the whole idea is about getting Earlier detection and less undetected outbreaks, for example So did you measure that also in in in a way if it's if The decisions were taken more swiftly if the response was given more swiftly So and at what level the decision is taking because routine Surveillance as you described it is going up to the district and then to the national level But I'm not clear at what level The decision will be triggered At what level the decision will be triggered by this data Thank you, and I was wondering right on the back Hello, David MSF Sweden innovation unit I was wondering a bit all the tools and systems that you are presenting I mean they are mainly for us to be able to have a better overview on To be better prepared to have better information on where to go and what to do. I was wondering Is there any thoughts about the next step on how the population themselves could benefit From all this information that we are gathering? I mean, we know that people are having more and more access to Mobile phones and technologies in the most remote areas in the world So I have a feeling sometimes that it's very interesting for us and it's very important But it's like we are looking from very high up Collecting what we need for our for our activities to be better But is there a future thinking on how people having mobile phones themselves How they could use or benefit from from all the Activities that we are doing So, yeah, open that up to the panels. We have Whether the data that was collected at the health facility level was used for decision making So, yeah, I think it's to you two at the end. I think So the in the PAP project it was really a very close Partnership with the MOH on a province level And the decision was made between the The supervisor of the Prefective Sanitaire or the MOH local office And the MSF office with the medical coordinator in particular So there was a lot of Exchange of information and communication that about what to do when there are certain cases, especially if it's it requires any Action immediate action. So this was something done on a continuous basis that to Report the data and that's why also the GIS Visual map has been developed by MSF as well. It was very helpful to really see on the spot What's happening in terms of cases and death? So I would say that it was managed case by case it depends on the types of diseases as well So this was done between the MOH local office in Mamberika D and the MSF Regarding the component of transmitting the data from the province to the capital city It's outside. It was not part of this project because MSF is not really about surveillance the action As a humanitarian organization is action oriented. So that's why the focus was always on the healthcare center to the Main office in the province We have to wrap up very shortly. So if you have a very quick response Yeah, well for us that was the main the main point not information for action and actually we were meeting every Monday Having a look at the dashboard the data that we have collected and seeing if that was any Exceedance and from there we will go to the map Identify where that was clustering and then try to follow up with the phone and Visiting if it was nearly so yeah, and the decision was always making at the distril level Okay, so unfortunately we're out of time, but I'd like to thank the panel members again for their presentations on their great discussions