PICCABI

We are developing a novel, groundbreaking technology to assist people with the need and desire to accurately track their diet to live a better life.
This technology will allow the collection of much-needed data on food consumption for organisations that need to run risk & exposure analyses to ensure only safe foods are placed onto the market.

We have recently finished a commercial feasibility case study during which we interviewed 18 companies including providers of food tracking systems, such as mobile apps, food producers and retailers, government and expert organisations like the Food Safety Authority of Ireland or Diabetes UK. All the companies showed interest in the tool and confirmed that it can solve two of their major problems: tracking people's diet in an accurate manner and collect data about food consumption over a long period of time.

Three of these companies have signed letters of intent to use the technology of PICCABI and implement it in their current offerings through API access licence agreements. They are also willing to collaborate on the development of the PICCABI tool by opening their user bases for early trials.

The Diet and Nutrition market is projected to witness the highest growth during the forecast period. The global weight loss and diet management market size was valued at $192.2 billion in 2019 and is projected to reach $295.3 billion by 2027, registering a CAGR of 7.0% from 2021 to 2027 according to a 2021 report by Allied Market Research. The growing rate of obesity and number of diseases associated with it primarily drives the market growth. Thus, the strong momentum behind the creation of PICCABI would resolve these concerns by making it easy for the user to grasp portion sizes. This trend has been further enhanced by the COVID-19 pandemic. According to World Economic Forum, the downloads of health & fitness apps has grown by 46% at the global level in 2020, with India experiencing a whopping 157% increase. In Q1 of 2020, there were 593 million app downloads.

The Total Available Market (TAM) is likely to be in the order of around billion users worldwide, but for conservative reasons we build our Serviceable Available Market (SAM) and Serviceable Obtainable Market (SOM) projections on the numbers that we can stand up to. Hence, we take as TAM the global number of users of MyFitnessPal and FitBit which sums up to 231 million users. We have the ambition to reach the global market, but we will start from the domestic one, namely Ireland and the UK, which represents 1% of the global population. This means that the SAM is in the order of 2 million users, but we will probably be able to reach 5% of them in the next 3 years bringing the initial SOM to 100,000 users.

Consuming a healthy diet throughout a person’s life helps to prevent weight gain and malnutrition in all its forms as well as a range of food-related diseases and conditions. Tracking the diet serves as a source of motivation. Many services allow you to do it but none automatically estimates the portion size in addition to food recognition. Currently, users of diet tracking technology must weigh each food if they want to accurately log them. Thus, people can shy away from the thought of food logging, because it is time-consuming and frustrating and, nowadays, people are always on the lookout for the quickest and easiest methods using modern technologies. This leads to low user retention rates and huge revenue losses for the providers of these services.
The lack of a tool capable to accurately estimate portion sizes makes the current food tracking process prone to significant errors. Studies have shown that people can under-report their food consumption by anything up to 50%. This can defeat the purpose of logging foods and generate more frustration in the users as they might not see the expected results in terms of weight loss and health improvement.
Moreover, accurate, long-term consumption data is vitally important for large corporations and government agencies, either for commercial market analysis, scientific research or for establishing regulations. They can use this data to ensure that their products fall within regulations by modelling exposure to food hazards on consumption data. However, this data currently is, at best, 2-years old and times change quickly. Recently, the European Community has funded the Richfields project which aimed at creating a consumer-data platform to collect information on food behaviours from different sources, including datasets curated by academia and corporations. This proves that this data is in high demand. Our tool will kill two birds with a stone by helping providers of food tracking services enhance their products and generating vital data for several organizations.

The envisioned product is a food-tracker system, with robust API access for other food tracking and fitness apps (like MyFitnessPal and FitBit, for example), which will have the following innovative features:
* A food recognition algorithm that can estimate the portion of the meal with a high degree of accuracy by leveraging cutting-edge computer vision and deep learning technologies.
* A guided, step-by-step process that allows the user to add information on food add-ons (such as sugar and salt) or personal ingredients (replacing cow milk with soy milk, for example).
* Users will be able to share their experience on social media and keep focusing on their health and fitness goals. This is to use the already common habit to take and share photos of meals on these networks and add a game-like layer to the diet-tracking experience.
* Fitness apps developers and social media companies will have an excellent plugin to enhance their products.
* The system will allow the creation of a food consumption dataset of unrivalled accuracy that can be made available to companies and institutions for research or marketing purposes.

This solution will specifically address the tedious task of manually weighing and recording consumed foods on a diet tracker. This leads to low user retention rates of food and fitness apps and the lack of accurate data on long-term food consumption for research and market analyses.

This is where PICCABI can change the current limitations of the diet and nutrition market. The SaaS solution goes beyond what is available at the moment that consists of several standalone nutrition apps, such as Asken Diet App (www.askendiet.com), Calorie Mama API (caloriemama.ai), Foodphone by Smartdiet, LLC (https://www.foodphone.biz/), FoodTracker (apps.apple.com/us/app/foodtracker-count-calories), Foodvisor (www.foodvisor.io/) and SNAQ (snaq.io), just to name a few.
However, the fitness and nutrition apps are not competitors, per se, instead, they are potential partners. Through leveraging our industry contacts and business development we will establish partnerships to gain the most exposure to the market.

The current market leaders have complaints about the low level of accuracy. This is a crucial problem for apps aimed at counting calories and nutrient intakes. We’ve carefully reviewed different nutrition apps and we are confident PICCABI offers superior features. None of these apps is capable of estimating the food portions in an accurate manner.
PICCABI is developing a novel technology that cannot easily be copied. The training of the computer vision models that will be used in the PICCABI tool requires many photos per food type. The collection of this data requires time, money and expertise as the photos must meet various requirements.

Fr om a business point of view, this is a great opportunity to attack a huge and growing market (health&fitness technology) from three different angles:
1) An integrable B2B SaaS component that can appeal to the existing major market players (such as FitBit, MyFitnessPal, Google Health) and other providers of nutrition apps. Potentially, anyone can benefit from adding this functionality to their product
2) A standalone mobile app that targets consumers and will help us to grow and collect even more data.
3) A data analytics platform wh ere private companies, researchers and regulatory bodies can access anonymized data and discover all sorts of interesting insights.

The main costs will be related to the salary of the staff, as PICCABI will require professional figures that are in high demand, such as experts in computer vision, cybersecurity and data analytics. Besides, we will need teams for marketing and customer service support and experts in commercial law and GDPR as the photos might be considered personal data.
Another important cost centre will be the technological infrastructure to run the computer vision models of the food tracking tool and store users' data.

We are considering two pricing strategies for licencing the PICCABI’s tool.
The first one will be to charge on a per-user basis. Our tool can be considered a premium feature. MyFitnessPal and FitBit charge their customers USD $10/€8 monthly for their premium features. As PICCABI will be integrated into another diet-tracker tool, we will not ask for the full price as it must be split with the tool provider. We assume a price-per-user of €4.
The second option is to charge a per-licence price which will also depend on the number of users of the tool provider. We assume the following price options per month:
1) Less than 50,000 users - €100,000
2) Between 50,000 and 100,000 users - €150,000

The price for customers with larger user bases will be determined via a negotiation process.

We will have a separate price list for accessing the food consumption dataset via a fully customised analytic platform. The price will be agreed on a per-customer basis, according to the degree of customisation required by each customer and the number of accounts. We assume an average price of €100,000 per customer per year.

The MVP produced with the funds will be used to pitch investors in the last quarter of 2022 with the scope to raise new investments in the first quarter of 2023. The target of the second funding round is at least €1,500,000 which will be spent to hire new team members and consolidate the company structure, develop the MVP into the beta version of the tool, launch it on the market and generate revenue. It is expected that the investors will make a return on their investment mainly from the following revenue streams:
1) Consumer App Integration - Almost all revenue (~€7 Million) will be reinvested and invested into the company to ensure market competitiveness throughout the first 5-years of PICCABI’s existence. The target of €7M is a realistic estimate based on the revenue of similar companies and our pricing assumptions and sales forecasts.The total re-investment figure is to be determined, but the founding team is aligned with continuous investment to ensure market competitiveness.
2) Corporate Sales of Data Access - Within 3 to 5 years, we aim to receive huge ongoing licence sales (>€12 Million) that will need very little development and a huge market to tap into. This means that by year 5, any early investors can receive their investments back with a healthy return. The leadership team is targeting an exit strategy within this timeline (approx year 5).

Money will be spent on the following tasks:
1) Data collection - Collect an initial critical set of training images that can support a sufficiently reliable model for the market.
2) Model algorithm development - The goal is to create an MVP of the food tracking tool. This task can be divided into the following subtasks:
a) Define model design and requirements
b) Implement reliable image segmentation into the model.
c) Make the weight estimation part of the model resilient to the picture angle.
d) Validate the model.
e) API strategy and development
f) Model auditing and certification
3) Cybersecurity - Hire experts and implement all the necessary security systems to protect users’ data and company sensitive information.
4) Data analytic platform - Develop a safe, reliable and quickly accessible database that can store enough data. This objective is the estimation of the required storage space. This estimation must consider the expected/needed growth in the number of photos we collect.
5) Legal, GDPR and business insurance costs
6) Stationary and materials - It includes laptops, printers and any other equipment necessary for the staff to carry out their jobs.
7) Marketing includes costs related to hosting and maintaining a website and other social media accounts, advertising, and business travels.

We will be offering a percentage of the company's shares. The percentage will depend upon the offer made by the investor.

The main threat to our business might come from big tech companies. If big database holders, like Google and Facebook, decide to move into the food-tracking market, they can quickly become competitors and drive us out of the market.

The current computer vision technologies and models have some limitations that might not be easily overcome. For instance, the models can be sensitive to the condition the photos are taken (for example under or over light exposure, presence of other objects that can trick the model).
However, we have run two exploratory studies in collaboration with computer vision experts of the Insight Centre (based in Dublin, Ireland) to test that the proposed food tracking tool can be developed with the existing AI technologies. The results were promising and we managed to reach a satisfactory degree of prediction accuracy, even for the weight estimation, with a small training dataset. The computer vision experts have also advised on possible solutions to overcome the model sensitivities to variations in the input data. These strategies are still to be implemented and tested.

We participated in the first phase of the New Frontiers acceleration program organised by Enterprise Ireland.
Enterprise Ireland has also provided us with the funds to run a commercial feasibility case study and two technology feasibility studies.

We are applying at various competitions, but we are still waiting for their responses.