Poseidon-AI

Intellectual Fish Farming with Feed Optimization and Water Quality Monitoring​ using sensing and AI/ML

Our customers are cooperatives and/or medium-large scale aquaculture farms. We have contacted to more than 300 of these targeted customers in SEA and also; 2 government organizations. 80% of them pointed out the main issue for them is the high feed cost and they are unsure about the amount of feed wastage. Further, 20% of them pointed out the high mortality rate due to various reasons and oblivious on how to control. we have preorder revenue and organizations such as United Nations has already ordered our approach.

Our customers are cooperatives and/or medium-large scale aquaculture farms. As mentioned, our customers are medium/large size farms and cooperatives which have an average of 50 to 100 ponds. Hence, we have a one time on boarding fee for our devices and based on the saving, we will charge them a subscription fee of 10% of saving. I am attaching our business model.

There are 3 elements in aquafeed: fish meal, fish oil and crops. Fish meal and fish oil are very expensive to be made in laboratory condition and hence, humankind uses the species from the oceans to extract these two elements. With overexploitation of oceans, the species are heavily depleted and thus, by sustainable use of aquafeed, we will manage to give more time to stocks in the oceans to recover as well as help farmers to reduce their costs. We use image processing as a feedback to calibrate our algorithms and save up to 20% of feeding costs.

Feeding consists 80% of any aquaculture farm. The 3 main elements in aquafeed are, fish meal, fish oil and crop but due to high cost of producing the first 2 elements in laboratories, we exploit them from waters around the world. Under this condition, we are using a combination of affordable hardware and intelligent software to optimize feeding and save farmers up to 20% of their feeding costs.
Around 90% of seafood we consume is produced in SEA & China. On the other hand, the climate warming is impacting this region faster than others and since the species are cold blooded animals; this will heavily impact the energy consumption of species causing higher feed consumption but lower yield for farmers. For this reason, we use our devices to monitor water and use the parameters in our algorithms to present the time of feeding, amount of feeding, growth and maturity of the species, saving up to 20% of farms’ feeding costs. The device also carries 3 cameras which gives live feedback in order to conduct image processing and increase the accuracy of our algorithms for every species, in different areas and under different water qualities. We are also aiming to use air and water temperature data gathered from our devices to help NASA further develop climate models for small waterbodies in remote areas in SEA and China.

eFishery (Indonesia): eFishery makes a feed dispenser and automate feeding. However, Poseidon optimizes feeding by taking advantage of environmental changes in ponds. The eFishery uses sonar systems to locate the biomass and estimate the total biomass while Poseidon, take an individual approach and analyze it to total number using our algorithms. eFishery, uses sonar and have no camera, while we are using 6 sensors and 3 cameras. eFishery devices need installation while ours doesn’t and finally, we train the farmers and labors by having our devices as remote experts on the ground while eFishery, automate feeding in a field that is very labor driven.
Aquabyte (Norway): Aquabyte uses underwater cameras to run image processing and machine learning for disease control and feed consumption rate. The method used by aquabyte is unique, but it can only be used in waters with high transparency. For this reason, the method is only limited to high value species like Salmon. Hence, aquabyte can enter to the markets in Norway and Chile. Additionally, the method is expensive and not affordable for farmers in SEA and China. On the other hand, our method is affordable and can be used for various species with different water quality and under various transparency levels.
Tidal: Tidal system is very much based on underwater image processing and sensing. The system needs installation in waterbodies and is very on sight labor driven. Additionally, the system can work best for waters with high transparency and hence, they are best fit for species such as Salmon. Poseidon, on the other hand, does not need any installation, most of our work is done on our cloud storage, we use image processing to train our algorithms and finally, we can provide our services for every species living in different waterbodies and with low water transparency level.
Innovasea: Innovasea makes water quality monitoring tools which can be sent and deployed in different waterbodies. However, due to high price and necessity to install, its not very suitable for species in SEA and China. Additionally, the monitoring factor that Innovasea is offering will just show the condition of the waters but will not add further value to improve farmers sustainable behavior.
Umitron: Umitron uses IoT devices to optimize feeding concentrating on fish population. Their devices need installation and a large team to operate them while image processing on the background is only for transparent waters. On the other hand, how price of the device limits their market to high value species and only with the support of governments which will reduce their expansion.

The know-how, massive data-pool and software algorithms with sensor fusion will be hard to replicate. For existing locked-in customers, any close competitors have to provide 10X better solution to overhaul the whole existing sensor and software infrastructure. We are still early in the market with 1000’s of existing contacts and customer with trust which builds over time. Further, we will be a one stop shop for smart aquaculture by defining the standards, fish feed models and proven technology.

We have customers in countries all our SEA and China. We received 50K Revenue prelaunch and currently, we are training our algorithms for 6 species receiving subscription fee.

We call ourselves the Netflix of agritech because of our unique business model. We have one time on boarding fee for our devices and monthly subscription fee which we receive from the amount of saving for each pond, in each farm, under different environment and for different species.

Our biggest cost is the device cost. Currently, we are getting pre-orders from our customers to be able to make the devices, but due to limited numbers (less than 30 devices); we are unable to reduce the cost of our devices. We are planning to use the investment amount to produce larger number of devices so we can reduce the cost of our devices. Additionally, we are aiming to train our algorithms for 6 species (catfish, redsnapper, grouper, tilapia, seabass and perch) and we can use this funding to hire image processors, software engineers and data analysts.

We will be offering around 20 to 25% of the company for the amount we are planning to raise.

Our devices use 4G LTE to pump up the data to our Clouds in AWS server located in Singapore. This means that in every country, we need to have simcards from every country and thus, we are negotiating with service providers in these countries to provide us with affordable and reliable simcards. Additionally, in some countries such as China, the data can't be pumped up to our Clouds due to great firewall of China. Hence, we needed to change our Clouds to Alibaba Clouds which have servers located in mainland China.

We received a pre-seed funding from HAX and ef. The amount was 180 K USD, as a convertible note for 17.6% of the company.

UNIDO Innovation Ideas and technologies Vs. COVID-19 and Beyond