- 2 months ago
Agenda AWANI speaks with Datuk Seri Isham Ishak from Ministry of Agriculture and Food Security (MAFS) and Bjarte Olsen, Managing Director, Global AI Village (GAIV), on the digital innovation and AI to future-proof Malaysia’s agriculture sector with RAKAN TANI, 9.00pm tonight.
Category
🗞
NewsTranscript
00:00Hello and welcome to Agenda Awani. I'm your host, Fahna Sheh.
00:09Now, around the world, artificial intelligence is changing industries and daily life.
00:14And agriculture is no exception.
00:16In Malaysia, the Rakan Tani Initiative is showing how AI can guide farmers and strengthen food security.
00:24Now, today we'll explore what this transformation means with Datuk Sri Isham, Secretary General of Ministry of Agriculture and Food Security,
00:33as well as Beate Olsen, Managing Director of Global AI Village.
00:37Thank you so much for joining us, gentlemen.
00:40Right. We want to dive straight into what Rakan Tani actually means to farmers here in Malaysia.
00:47But let's give the overall picture here first, right?
00:50Datuk Sri, we're entering the new digitalization era with AI transforming industries everywhere.
00:56So for Malaysia, how is this shift changing the way we think about agriculture and the sector as a whole?
01:03Thank you, Fahna.
01:05Well, before we begin, we have to understand that in the world today, we are faced with a few megatrend challenges.
01:13Number one is on climate change.
01:16That's affecting the farmers and everyone else who's living in the world.
01:20But mostly focusing on farmers, climate change is the most difficult challenge that they are facing today.
01:26Secondly is the fact that many of our farmers in the country are quite old, ageing farmers.
01:34Average age between 60 to 63 years old.
01:37Thirdly, less participants by the youth to be active and interested in agriculture.
01:45So now with the introduction of AI and technology in agriculture, things are set to change for the better, of course.
01:54So with this timely introduction of AI and support by government, project Rakan Tani, especially the joint collaboration between Bernas and GIV, we have come out with an apps.
02:11And these apps are based on the WhatsApp feature that everyone already has been using every day on their handphone.
02:19So it's like having a new car, a car you don't have to teach them to drive again, but with new features on the car.
02:26So this Rakan Tani project is a new initiative, proactive initiative by Bernas together with GIV to introduce AI, especially to farmers.
02:37And as I mentioned today, climate change, old age farmers, are they ready to accept something that's technologically advanced to help them, assist them in everyday agriculture work.
02:49So with this WhatsApp based technology, we think that the paddy farmers today facing many, many challenges will be able to be assisted more precisely, technology driven with data to support whatever actions that they are taking so that they lower the risk of their everyday farming activities and their successful rate is much more better.
03:12And of course what we intend to do with this AI technology is to enhance and ensure that their productivity when they do paddy farming is increased and also more importantly is to get the younger farmers or younger generation to be interested in agriculture.
03:32So those are challenges that we face today and with the advent of AI through this WhatsApp based technology, we are confident that this proactive initiative will definitely provide the impetus and boost in the paddy farming activities in the country because paddy farming, paddy of course, is one of our major staple food.
03:56We cannot live without it. We cannot live without it. We have to have it. And technology will definitely ensure sustainability in this sector for many, many years to come.
04:07Yeah, I mean, as a Malaysian, I can concur that we cannot live without rice here. So yeah. But I want to talk about, you know, you mentioned just now a keyword, which is readiness, right?
04:16Because, yeah, because, yeah, like you said, it's driving a new car, there's already, the car is already there. But now there's this new feature that people have to adapt to it, right?
04:26So, from the ministry's perspective here, so what concrete steps are being taken to make this digital shift a reality on the ground for these paddy farmers, right? So how are these initiatives being translated and to help farmers to be ready to adopt AI?
04:44Okay. Well, AI, the backbone of AI is always data. And behind that data is an AI engine that helps create this information realm for the paddy farmers, especially on the system.
05:00So from the ministry's perspective, we have collected many, many years of data into the system. And this hyper-personalized system actually customizes every plot of land to that specific farmer.
05:15So data-driven technology-based AI will really change the landscape for the paddy farmers, farmers especially.
05:25So paddy farmers traditionally have been practicing traditional farming. Some of them keep all this information and knowledge in their head.
05:33But sometimes some of them write down in small notebooks. And these are collected and aggregated throughout the years and put into this database system.
05:41And the AI engine now from GIV generates that very critical information for the paddy farmers whenever they are doing their paddy farming.
05:52From harvesting, all the way to tilling and starting the cycle again. So throughout the cycle, the farmers are assisted via this system based on many, many years of data.
06:05And hopefully this provides confidence for them to make sure the sustainability of their paddy farming continuously improve over the years via the AI engine that's supporting new data that's coming in.
06:20Personalized data, personalized plot where they farm, weather in that area, the soil that they are using.
06:28So many factors are now being supported by data and providing confidence that we hope will ensure sustainability of the farming for many, many years to come.
06:40So I guess that's where you come in, Beata, which is, you know, with Rakan Tani AI.
06:45So could you walk us through what it actually does and why it marks such an important milestone in Malaysia's food security efforts here and for the paddy farmers here in Malaysia?
06:55Okay, thank you. Yeah, Rakan Tani, as mentioned by Dr. Suri here, is meant for the farmers.
07:03So it's supposed to be a companion for the farmers, supposed to be friendly, easy to interact with.
07:08And it's for that purpose driven by a WhatsApp channel. So they don't have to install a new app, familiarize themselves with the new technology really.
07:17They are meant to simply add a phone number to their contact list and start communicating.
07:21So what Rakan Tani does, in essence, is to democratize a lot of this information.
07:28So some of this information might be written down, but that means it's not necessarily circulated.
07:34So that means the best practices are not spread. So by taking all this information and democratizing it,
07:40together with research information from universities, from research institutes, etc.
07:45and from Bernas, who spend a lot of time on this, we are able to disseminate and send out this information to farmers as and when they need it.
07:55So that's one part of it. But what is quite crucial and I think special of this solution compared to just a chat bot, if you like,
08:03where you can ask questions and get answers, is that we are coupling this with the actual rice schedule, the JKL for the farmers.
08:13So that the most innovative part here, I would say, is that we are, through a natural dialogue with the farmers via WhatsApp,
08:20whether it's text or voice, we can try and understand where in the schedule the farmer is.
08:26So we know what are the upcoming activities, what is needed to perform these activities, what kind of inputs are needed,
08:33whether it's seeds, whether it's fertilizers, pesticides, they can know this upfront and start preparing and have it ready.
08:40So when they're going to perform an activity, it's already available at their hands.
08:45And then the information comes in as well and says, oh, this is a new type of pest detected.
08:51The government has maybe approved a new type of fertilizer, a new type of pesticide,
08:55which maybe they're not familiar with.
08:58So how to apply it, should they do it in the morning, should they spray it, what are the ratios, how should they mix it.
09:04All this information can be given to them without them having to study a lot about it.
09:08So this coupling of information with the schedule, I think it's very important.
09:14So you can kind of think of Wakantani as a mentor, a co-pilot that is with you all the time
09:20and adapted to the individual farmers, but precise location.
09:25So it's hyper-personalizing of the schedule to adapt to weather, to pest outbreaks, to any delay you might face as well.
09:35So there's a lot of changes to weather patterns.
09:37You don't really have the distinct dry spell, wet season so much anymore.
09:41You have erratic weather, which can be hard for the farmers to adapt to on a really rapid note.
09:46So the idea here is to try and predict this through having good integrations with weather information,
09:55satellite information, et cetera, to be able to provide additional layer of information to the farmers.
10:01However, we are also cognizant that it can be a bit difficult for farmers to adopt a lot of new technology.
10:09So we don't want to overwhelm them with a lot of devices, et cetera, that they have to install and set up in their farms.
10:16But we can, of course, integrate with them should they have these things in place already.
10:21Right. I think that goes to my next question, right?
10:37Because we talk about the practicality of the app or, you know, the co-pilot WhatsApp for Farmers, right?
10:45How exactly would it be integrated fully with farmers?
10:50Because farmers face real challenges. You mentioned just now, you know, pests, unpredictable weather, for example.
10:55So how exactly does Rakan Thani actually help them deal with these day-to-day issues?
11:01If you don't mind sharing, you know, some examples so we can sort of paint a better picture on how farmers can actually utilize this
11:07and how are they going to be trained to use this?
11:10Yeah. So it could be from the early stages when they're preparing the land.
11:16If they do, for instance, a soil test and they check a pH value.
11:20And certain parts of the country have different levels.
11:24So some places they are okay to have slightly more sour, slightly more acidic.
11:28And you can ask Rakan Thani, is this okay? Is this within the limits?
11:33Or what should I do if it's not, right?
11:35So then you would get automated examples and recommendations for that.
11:40Or it could be when it comes to water, it could be that, of course, rice farming is very dependent on water and access to water.
11:48So schedules from organizations that control water could be integrated into the system.
11:55So they know exactly when they have access to irrigation, etc.
11:59Other things, like I mentioned already, is within pest and fertilizers.
12:05A lot of this can be obtained by the farmers, so from special outlets.
12:13So availability in these outlets could be one thing where we are trying to integrate with those as well.
12:19So we have a bit more control on the availability and that they have access to what they need for their particular farming activity.
12:26One thing that we are looking at incorporating now also is, as we deploy this to more and more farmers,
12:32you will have a community of farmers, a bit like waste from your traffic rights.
12:37Your waste will be able to tell you where there is a congestion.
12:40Similarly, if we detect that some farmers are reporting a particular bug or pest,
12:45if we see that at a number of farmers in the same vicinity,
12:49we can kind of decide that, okay, here seems to be some kind of outbreak.
12:53So we can give early warnings to farmers, so they are prepared in advance for that.
12:59Right. I mean, it seems that this co-pilot or this app would really help strengthen the farmers' capability
13:08in harnessing their already given knowledge on these kind of things.
13:12But I do hope there is a lot of awareness and a lot of training involved in order for them to adopt this.
13:18And I think Datuk Sri will address this later.
13:21But let's look back at what, you know, this app or this assistance will do in terms of opening up the rice sector, right?
13:30You know, we in Malaysia are moving forward with AI for the right yet,
13:37in terms of, you know, the madani economy and making sure that, you know, nobody gets left behind.
13:43So looking ahead, what opportunities could you see with regards to this advancement that could open up the rice industry industry?
13:53Well, you mentioned about the madani agenda and no farmers left behind or no communities left behind.
14:02I think that's the objective of what we are doing here is to find the right formula to optimize the resources that the paddy farmers have.
14:12Optimizing in terms of the land that they have, water resources that they have, fertilizers that the government sometimes subsidizes, and many others.
14:22Because the cost to the farmers if there is a loss to the crops is huge.
14:28And if they incur losses, sometimes they lose hope.
14:31When they lose hope, then we lose our paddy farmers and we lose the rice itself.
14:36So we have to make it right.
14:38So it is important to make sure that the AI system that we provide must be simple, easily to be used, is sustainable,
14:49there's always continuous improvement.
14:51And what Vialty mentioned just now is the expert technical companion.
14:57There are more than 100,000 paddy farmers registered in Malaysia.
15:01And I mean, there's a few of us, few of the technical officers from Jabatan Pertanian who's there to help and support them.
15:09So with this AI technology provided to personally become the expert companion to the paddy farmers,
15:19we think that the optimization effort will be able to help them be confident that whatever they do is correct,
15:27is accurate, timely in terms of timing.
15:31And they are more confident now that whatever they plan, insya Allah, the yield that they will be getting is more predictable for them.
15:40So predictability is very important here.
15:42Optimization, I mentioned there's now a keyword to ensure this.
15:47So with all this, now what we need to do is try to share this with other farmers to say that,
15:54hey, this technology works, it has been proven, so you try it and see how the system could double or increase your yield now.
16:03Because we have a technical companion to be with you all the time,
16:07personalized to your type of land, to your area, acreage that you are planting,
16:13and also the type of variety, paddy variety that you are planting.
16:17So you should be confident, you should know that this helps optimize your opportunity now to ensure a higher yield and productivity in whatever you do in the paddy farming.
16:31Yeah.
16:32Yeah, I mean, I just think it's important to emphasize that we are not necessarily changing so much what's already in place.
16:41There's been a lot of good initiatives from the government side, from organizations like Bernhardt's with a large scale farming practices.
16:47However, they are dependent on having staff there, a field coordinator in place.
16:53So what we are trying to do is to help them to make a virtual field coordinator so that more farmers can get the benefit of having this advice.
17:01So I want to assure you, yes, you are actually doing the right thing.
17:04What you're doing is good and you are going to get increased yields from your practices.
17:10I mean, it's a sort of validation that you get from, you know, AI, right?
17:14For example, with most of this stuff you already know, but it's just the sort of validation you need to move forward, right?
17:22To ensure that, okay, I can move forward with this best practices.
17:27It's already been sort of confirmed with this AI chatbot.
17:31And, you know, if I need more information, I can always ask more and I can always explore and find out more.
17:37I think it adds that layer of advice, tips and tricks for them to sort of navigate, you know, the changing weather, the past, the type of soil, any new advancement, any new trends.
17:50So, yeah, I think all around it's a very good assistance that you can have in your pocket from day to day, right?
17:57But I think, okay, we have this technology and it's great and it has a lot of tools that farmers can use.
18:05But I think if you talk about, you know, technology adoption here, it's never without challenges, right?
18:11So, what are the main barriers actually when it comes to farmers embracing AI tools like this?
18:18You know, trust is a big issue with a lot of people who are not familiar with this kind of technology.
18:23So, how do we tackle this and what lessons can we take if you want to apply this to a wider scale?
18:32Maybe, Biata, you can take this.
18:34I think you're right.
18:36All technology, introduction of new technologies has some issues.
18:41So, one, we are trying to use a very familiar medium for interaction with WhatsApp.
18:45That's number one.
18:47We also, as I already mentioned, we try to limit the need for farmers to actually buy any kind of equipment that is extra.
18:55So, it's low capital front cost.
18:59And then accuracy is really important.
19:01So, we are doing a lot of trials with Bernas to get feedback from the farmers on the ground.
19:06We're trying to build in mechanisms where they can very easily actually conversationally give feedback as well.
19:12Because it's difficult to get farmers or anyone for that matter to actually fill in a form and say, oh, this is my feedback.
19:17Right.
19:18So, if they can in their WhatsApp conversation with Rakantani say that, hey, I'm not so happy with this.
19:24Or maybe why are you answering this?
19:26We can pick this up.
19:27So, the AI can actually pick up that this is actually feedback for us.
19:31And we can put that aside and we can run this through feedback loops, etc.
19:35To improve, constantly improve the product.
19:39Another thing is, of course, that they might not know exactly how to use it.
19:45Right.
19:46So, typically with AI in general, a lot of people are using it like they would use Google.
19:50So, they would say, oh, I want to buy 50 kilos of a product X somewhere.
19:55How much does that cost?
19:57This is very difficult for Rakantani to answer because it doesn't necessarily have all this and current updated information.
20:03Right.
20:04So, this is where we are trying to work with both the government, with organizations, both for the farmers, but also on the private sector to see how can we integrate more of these kind of inventory like sources.
20:18So, we can get very accurate responses on these things as well.
20:23Giving them additional information, like giving them weather updates that they currently have to look up.
20:30I know the farmers, when I spoke to them, said, oh, we are really, really careful about the weather.
20:34We check the weather forecast four times, five times, six times a day.
20:37Now, they will get that automatically for their precise location in addition to having it incorporated into the advice.
20:44So, we are, of course, looking at doing this in a larger scale, maybe different countries as well.
20:52So, it's important for us that we can fully test it here, get the best practices, get the feedback and the buy-in from the farmers.
21:00Because the farmers are really the users here.
21:02They are the ones that are making sure this is a good product.
21:06Because we are also dependent on the interactivity with them.
21:09If they just ask a question every once a month, it's very difficult to get an updated information source as well.
21:30So, two questions here.
21:32Because I'm curious to know, because I'm actually impressed by this app and how it would help.
21:38But in terms of language barriers, right, will there be any sort of, you know, review?
21:46Because farmers sometimes they use sort of their own languages, their own jargons when it comes to, you know, farming practices.
21:52I don't know most of them, but I'm pretty sure they are.
21:56And would the AI be able to adapt to this?
21:59And also, like you said, you know, a lot of these people treat AI like Google, right?
22:04They ask one question and that's it.
22:06Will they sort of know that this tool needs to be trained?
22:11Yeah.
22:12On the language side, yes, we are trialing a lot of different components to be able to detect the different dialects, etc.
22:22So, we did successfully trial Hokkien mixed with Clantanese.
22:27Wow.
22:28Which was quite interesting that it could actually understand a lot of that.
22:31Right.
22:32And that was quite impressive.
22:34But there are a lot of local initiatives.
22:37I think this is an important point as well.
22:39We are not trying to replace anyone.
22:41So, initiatives currently done by the Ministry of Agriculture, Food Security, initiatives done by research universities and the private sector.
22:49We want to collaborate with them.
22:51So, when it looks at there's not a lot of new language models being launched by different companies in Malaysia.
22:57So, there is a lot of effort happening on that side.
22:59So, when it comes to language detection, we are not trying to reinvent the wheel there.
23:03We are trying to use other companies and their expertise on that side in Malaysia.
23:08So, I think that one we are getting pretty good at.
23:12Okay.
23:13When it comes to the other point where they need to understand that this model needs input from the farmers.
23:21It's all about making this a very friendly conversation with them.
23:25And so far, they seem to be quite interested.
23:27You know, they feel that they can kind of get responses that are not just question answer, question answer.
23:33They will be, oh yeah, you are doing well.
23:35And they will remember.
23:36There's a memory function kind of there.
23:38So, it remembers like, okay, you said like a month ago that you are doing this.
23:42Have you actually followed up on this?
23:43Right.
23:44And, okay, you are doing great now, you know, and you are on your way to a great success.
23:48So, hopefully your yield now is going to be better and then you can take some leave after this, you know.
23:52It would be very good to hear on a daily basis.
23:55Exactly.
23:56That kind of encouragement, right?
23:57Okay.
23:58So, moving forward here, if you want to, you know, we talked about scaling up just now.
24:02And I think part and parcel of how we want to scale up is collaboration, right?
24:07And this project is a good example as a collaboration between, you know, these two agencies.
24:12So, from your perspective, let's start with you, Dato.
24:15How important are these, you know, sort of partnership in driving AI adoption here in Malaysia
24:20and amplifying it better to benefit today's society?
24:23Okay.
24:24So, speaking about technology for farmers, petty farmers especially, two words come to mind.
24:29One is adopt and the other one is adapt.
24:32So, adopting is the challenge that we face here because of the age differences between
24:37different, different petty farmers.
24:39For the younger generation, it's easy for them to adopt and also adapt.
24:44But for the older generations getting or to convince them to adopt, then later adapt is another challenge.
24:50So, we have had many, many types of promotional programs to try and encourage the older petty farmers
24:59to come on board.
25:00So, in the program program Madani Rakyat and all those major events that we have, we try to introduce
25:07by having them feel, feel the system, feel the technology and how it can help them actually.
25:14So, some of them are a bit scared, although they have their smartphones and some of them still have the older version phones.
25:20But those with the smartphones, we try to assist them via, we have simulators to try, for them to try it out, our programs.
25:30And every time when we have programs, the technology section is always a big attraction.
25:36Right.
25:37They love to come and see what new technologies that they can adopt.
25:40Then later on, perhaps, if they're convinced, then they can adapt.
25:44So, via the simulators, this is where we take the opportunity to understand better and further about
25:50what do they want from the system.
25:52So, what Beate mentioned just now about the different dialects in the system, that's very useful.
25:56But in the WhatsApp system currently in this first version, it's only what you type in.
26:02Right.
26:03And you can also voice, record whatever your questions are.
26:06But later on, I think when we are able to customise these questions and difficult questions by the farmers,
26:14in terms of them taking photos of how the paddy condition is like.
26:20Yeah, how it looks like, yeah.
26:21Later on, I hope there will be a feature, add-on, maybe a version 2, where the paddy farmers can now take a photo
26:27of their crops and understand better how healthy is that crop is.
26:32What are the fertilizers you require?
26:34Is there sufficient water or not?
26:36And whether there's a pesticide currently being invading their farmers.
26:41So, these additional features will definitely provide optimization to this system that you mentioned.
26:48So, the challenge that we have now is trying to get more people to understand what technology and how it can help them.
26:55And especially with this WhatsApp based technology, AI, I think this is a good start for us.
27:01And Beate mentioned going to other countries.
27:03So, we are very proud to say that this is a first mover advantage for Malaysia
27:07because I don't think there's another system that's similar to what we have.
27:11So, definitely we see that this will really provide a good start for our paddy farmers in terms of meeting those challenges.
27:17I mentioned about those megatrend challenges that we are facing today.
27:22Right.
27:23I think that pretty much wraps up our conversation on Rakan Tani and how it would benefit the overall sector.
27:30And I'm pretty sure with regards to how it's already, how the Ministry and Global Air Village is already trying to sort of look into, you know, the older generation.
27:41I'm pretty sure there's also a problem of younger generation going into the agriculture sector.
27:45And I'm pretty sure, you know, with the added advancement and added technology, that would be, you know, somehow soon realized as well.
27:52Thank you so much, Datuk Seri and Beate for your time.
27:56Now, we've heard tonight that Rakan Tani is more than just technology.
28:01It's a step forward to empower farmers and support communities and securing Malaysia's food security.
28:07Now, with the right partnership, like you mentioned just now, with Global Air Village as well as the Ministry as well,
28:13you know, AI can help agriculture become more productive, sustainable and as well as resilient.
28:18Thank you so much, both of you.
28:20Thank you for watching. I'm Farah Nashe and this is Ajinda Awanyi. Bye!
Be the first to comment