Two Mechatronic Engineering students at Jomo Kenyatta University of Agriculture and Technology Kenneth Kioria and Michael Mwaisakenyi were honoured by President Uhuru Kenyatta on this year’s Mashujaa Day for their invention dubbed ‘Weeding Robot’.
The Smart Harvest and Technology caught up with them to share insights on what inspired the technology and how it will change the fortunes of small holder farmers.
President Uhuru Kenyatta honoured you as heroes during the Mashujaa Day celebrations, how did that make you feel?
Kenneth & Michael: We were very excited that His Excellency President Uhuru Kenyatta honoured us. We did not expect to be acknowledged by him during his Mashujaa day speech. It is an honour to be regarded as a shujaa especially at a young age and we hope to inspire Kenyans especially the youth to come up with great innovations.
What inspired the innovation?
Michael: It was a response to a need I saw first with my grandfather and upon further investigation, realised every farmer is facing the problem of weed control. The weeding robot seeks to solve this issue in a sustainable manner.
With your unique knowledge, strengths and skills set, what unique role did each of you play in this invention?
Kenneth: Since we are both Mechatronic Engineers our skill sets are similar. We each designed what we had in mind, compared our designs and ended up combining both to get best results. We both wired the robot, programmed various aspects of the robot and did vigorous testing to make sure everything was working.
How exactly does the automated weeding robot work?
Michael: It is an automatic weeding robot that moves through the farm by itself through the crop rows and it has a tool that it drags to remove the weeds in between the crop rows. It has a robotic arm to remove the weeds along the crop rows. It uses a camera among other sensors and artificial intelligence to move through the farm and remove weeds. The artificial intelligence is trained using various images of the crop in our case, cabbage. Thanks to cloud platform services such as Microsoft Azure, we were able to train the artificial intelligence model in less than three hours.
What powers the robot and can it use affordable energy sources?
Kenneth: The power source of the robot is a Lithium Polymer LiPo (LiPo) battery of 8,000 mAh. We have four of those with capacity to run the whole day if need be. The batteries are also rechargeable and recharging takes around an hour.
What is the robot’s maintenance needs?
Michael: The mechanical components of the weeding robot are made to require little maintenance. For example, the metal parts are made of aluminum to avoid rusting. The motors we use are water resistant and rarely need replacement. The good thing is the customers don’t need to worry about the maintenance or the cost of the robot because we plan on doing farming as a service model. This keep the costs very low for the farmer to weed their farm while increasing reliability.
What are some of the unique aspects of the technology?
Michael: Currently, using herbicides for crop disqualifies the produce in the international market. The fact that the weeding robot removes weeds from farms without the use of herbicides is a huge boost to any farmer targeting export market. The small holder farmer should celebrate the fact that the robot does not have consumables and therefore allow us to sell the service at a very competitive price to the alternative herbicides.
Weeds are a big headache for most small holder farmers who are forced to use expensive and toxic herbicides, what are some of the tangible benefits crop farmers stand to gain by using it?
Kenneth: Given our business model which is farming as a service, farmers do not have to buy the robot and worry about maintenance, faults etc. Instead they can hire us and we do the weeding on their farms. This removes a lot of burden of getting and maintaining the robot, it is affordable and low risk.
Michael: Long exposure to herbicides has been found to increase the probability of getting cancer. Cancer being a big health risk caused by the primary ingredient called glyphosate in the herbicides, the weeding robot eliminates this risk while allowing the small holder’s produce to be able to enter the international market. The model we have allows the farmers to be able to afford these services instead of losing up to 60 per cent of their yields to weeds which ensures they have almost twice the produce they used to get for the same effort or less.
What stage of invention is it at and when do we expect to see it in the markets and expos?
Kenneth: We are now gearing towards the pilot phase where we plan to have seven robots and 10 farmers. From our calculations, we have seen seven robots have the capacity to weed an entire 5-acre farm in around eight hours. So, we are look for investors to fund the pilot phase for a successful run. Once we are done with the pilot phase and have done sufficient testing and improvements, we can start offering our services. We do not plan on selling the robot but on offering our services as we see it to more affordable and sustainable to the small-scale farmers. We have participated in JKUAT Tech Expo where we able to interact with a lot of people and bring awareness of the innovation we have made.
What are some of the .challenges you faced when developing this prototype?
Kennedy: The major challenges we encountered were acquiring of resources as most of the items were not available in Kenya. We had to import most of the materials especially the electronic components. They also took quite a long time to arrive, around two months since we ordered. Also, getting the capital to build this robot was a challenge. We had to ask our parents, dig into our own pockets and even plough back the funds we won from the Imagine Cup. It was and is an expensive process but it was all worth it in the end.
Some partners have expressed interest to help you commercialise the idea, tell us more…?
Michael: The weeding robot has attracted quite a lot of interest even from the imagine cup regional finals where we were getting very positive reviews. We had been approached right after that with someone having a project on smart farming. Farmers have approached us quite a number of times to buy one. The problem arises in the engineering process that the weeding robot must go through in order to be a product that is ready for the market.
We want our product to be of the highest quality and that can only be achieved and guaranteed when we have completed the pilot phase. We have run the numbers and the weeding robot is an extremely valuable investment and we have a good strategy mapped out. We don’t want to compromise on quality because we want to get to market and make money quickly.
What message do you have to your fellow budding innovators who are struggling to launch great innovations?
Kenneth: For those who haven’t started working on their idea, I urge them to start. Feedback on your innovations is one of the important things when you’re starting out. They shouldn’t spend a lot of time analysing the markets, whether they will fail or not. They should start with they have, build a Minimum Viable Product as soon as possible and see how people receive it. Once they have feedback, they can keep upgrading the innovation.
The support comes later but not before you start as now people including investors, see the viability of your idea. For those who have started, I urge them to continue with what they are doing, the obstacles and challenges will always be there. It does get easier in other ways but harder in others.
10 years from now, where do you see this technology?
Kenneth: I see it helping small scale farmers not just in Kenya but in Africa. We plan on taking on the African landscape as they are around 33 million small scale farms on the continent as small-scale farms make up 80 per cent of all farms in Africa. So, there is a huge market for our innovation and great chance to help many African farmers get better crop yields in an eco-friendly fashion.
In ten years, kids will be growing up knowing weeding robots as a common tool for the shamba.