We develop a complete system of detection of new species of flora and fauna, this system is called "Project Itza" and is composed of three parts, the first one is the creation of a robot whose design of a spider and an octocopter will allow you to overcome any obstacle that arises or access remote places. The robot has two cameras that will take pictures of its environment and with the use of artificial intelligence will compare the images obtained with those already existing in the NASA database, so that it can identify all the objects that surround it.
Once that analysis is complete, the photographs will be sent to our website along with the species name, percentage of coincidence, the location where it was taken, date and time. If a photograph has a percentage between 50% to 60% coincidence it will have to go through a second more specific analysis where if the resulting match percentage is 0% to 20% it will be classified as a possible new species and an alert will be sent to the scientist in charge through the website , so that it can then do the corresponding analyses.
The second part of the project is the design and implementation of strategically distributed power stations in the area the robot is exploring, the function of these stations is to provide charge to the robots when they run out of battery.
Robots will automatically be alerted when their battery is low, so at that moment they finish their journey and head to the nearest station. The goal of distributing charging stations is that robots do not have to return to the location of the scientist who owns them, achieving a broader search.
The third is the development of a website that will allow the scientific community to have access to the robot information in real time, as well as generate analysis of the interaction between species, their abundance, movement, distribution and their physiology. Through the website the researcher must assign the route (with coordinates) that each robot must travel and if he has identified a possible new species he can load that image to the robot to be his search objective.
Our website can be accessed with the URL: https://dev.d3m6bo6bxdky1h.amplifyapp.com/ and in the future could also serve as a means of disclosing NASA results and other research.
This system is of great importance, aid and impact because it is estimated that there are approximately 8,8 million and approximately 1,9 million are known. With this project we look forward to supporting the scientific community in the automation and efficiency in the search or recognition of new species. We also hope that our system can provide the necessary support to identify in a timely manner changes in the behavior of these species and thus stop or decrease the loss of biodiversity on Earth, as one of the latest studies conducted by "Live Planet" detected an overall average decrease of 68% of the nearly 21,000 populations studied from mammals , birds, amphibians, reptiles and fish.
We are a multidisciplinary team, which is composed of students in Computer Engineering and Mechatronics Engineering. We choose this challenge because we are young entrepreneurs, proactive and with innovative ideas, we like to test our skills and knowledge to implement projects with innovative technology that is focused on serving society, contributing to generate positive change in Mexico and the world.
We believe that this challenge is one of a kind and of great importance because it tries to know more about the planet on which we live, as well as to resolve in timely manner alterations in the biodiversity of this. Human beings must understand what our place is on Earth since the well-being of us is integrated into the welfare of all animals and the environment.
The focus of our project is to support researchers to find new species and to continuously observe species behavior, with tools that are easy to use and perform excellent work.
With our robots, scientists will be able to study places that are normally difficult to access or that could be life-threatening. Because our robots are autonomous and have artificial intelligence they can also be used to track endangered species and thus be able to propose a better solution.
For software development we use the following:
- Amazon Technology: AWS, Lambda, DynamoDB, S3 Bucket, AWS Amplify
- Image Recognition: Python, Keras, Panda, OpenCV, Boto3, Tensor Flow, Tensor Flow Lite
- Website: HTML5, CSS3, Bootstrap 4, JavaScript
For hardware design we use the following:
- Robot design: Solidworks
-Raspberry Pi, Pi Camera
For this hackathon, our achievements are as described below:
The problems were the implementation of accurate models because of the lack of specialized computer equipment, and for lack of time, we were unable to finish centralizing NASA data on our website.
Biodiversity information and ecological prediction were used, as well as their images that have been captured through cameras of different animals, the data we use as inspiration as we notice that they have a constant update on species, but what we notice is that more monitoring of such data is lacking.