1. General

Category

Category

Other

Please describe Other

personal safety

2. Project Details

Company or Institution

EvoAId Ltd.

Project

EvoAId Guarding Beats

General description of the AI solution

EvoAId Guarding Beats: AI based gesture controlled safety application with fall recognition
The technology allows users to trigger alert with a specific hand or leg gesture, and by fall even if
the smartphone is in the pocket or bag. As a result help can be called in Real-time that can save lives.

Website

https://www.evoaid.com/en/home/

Organisation

EvoAId Kft.

3. Aspects

Excellence and Scientific Quality: Please detail the improvements made by the nominee or the nominees’ team or yourself if your applying for the award, and why they have been a success.

Type of AI: deep neural network based time series classification
Sub type of AI: sensors based human activity detection

Worldwide unique safety application, with proven recognition model which is based on thousands of hours of human motion and tens of thousands of alerting gestures from hundreds of users. The recognition model is going to be trained continuously but thanks to the many years research and development and uniquely collected and wide range of training data the model has high-end generalization ability. Concerning false positive and false negative outcomes are minimized and normally do not occure.

USP: touchless alert; controlling gestures: leg, or hand; Fall recognition (based on fall, instead of lack of movement); no need for wearables, only smartphone required

Intellectual property: System for detecting a signal body gesture and method for training the system (PCT/HU2018/000039 pending in the USA and in the EU)

TRL: working application, already launched in Hungary via license partners

Scaling of impact to SDGs: Please detail how many citizens/communities and/or researchers/businesses this has had or can have a positive impact on, including particular groups where applicable and to what extent.

Safety is one of the most basic human needs. Rising concern regarding the security of children and women enhanced the demand of safety applications. The possession of Real-time alerting ability can increase the sense of safety. In case of sexual or physical violence an activated app can even warn off the attacker.

Due to the fact that the app doesn't require additional gadget/wearable we burden the environment less.

Scaling of AI solution: Please detail what proof of concept or implementations can you show now in terms of its efficacy and how the solution can be scaled to provide a global impact ad how realistic that scaling is.

There are applications with panic button and wearables in connection with application. In the event of attack, robbery or even health related emergencies there is no time to push a button in the app or even reach for the phone. Wearable technologies offer easier activation, but the users have to adapt to it, and their high costs hinder their market growth. There are voice recognition based safety apps as well, but those can't be applied by deaf or hard of hearing users.

Market research was conducted by our existing and potential license partners and more than 90% of the respondents considered our solution to be most feasible in emergency over other solutions. Also, more than 90% indicated, that our solution would increase their sense of safety, and they would recommend it to other family members.

We created an app which requires smartphone only and through AI approach and the applied process it provides working and universal alerting ability for everyone.

Network effect: our solution has enduced a spin off project which is realized in an international consortium and addresses elderly after stroke with aphasie or epilepsy http://www.aal-europe.eu/projects/aph-alarm/

Early adopters: we are focusing on B2B and have 4 license partners (security companies)in Hungary, who provide security service with the app

Related award: HTE Fekete László innovation Award – 2020 https://tekdeeps.com/an-infocommunication-innovation-is-worth-huf-1-million-again-this-year-technocrat/

Ethical aspect: Please detail the way the solution addresses any of the main ethical aspects, including trustworthiness, bias, gender issues, etc.

The application works regardless age, gender, skin color, origin and religion. The app is accessible, offers equal opportunities, we even made sure during the development to involve people with disability especially deaf or hard of hearing users.
The app complies with all the applicable laws, app stores regulations, applies transparent general terms and conditions, privacy policy. User’s permissions are transparent and held to the minimum. As developer we do not store and handle personal data except those sensor data which are anonymized and are required for the continuous training of the recognition model.
The AI based recognition model has high end generalization ability and robust enough for wide range of users.