Computer Robot

ORION

Automated Scientific Research

Robot

Orion is a tool created for the purpose of Automated Scientific Research. It has been written and developed in Python. Orion’s purpose is to help analysts and researchers reduce the time required to evaluate and summarize the vast amount of information (and constantly growing) related to the advances in technology, science, and engineering.

At its core, Orion is powered by a database of Open Access resources and Natural Language Processing tools based on Deep Learning algorithms. Only documents and databases licensed under CC-BY 4, CC-BY 3, CC0, CC0 1.0 Universal or similar licenses are considered. This guarantees our customers a frictionless/easier access to the information. Our current database is composed of 30,000 open access documents from top ranked publishers such as the IEEE Open and Springer Open, among others. Orion’s database includes over 400,000 images and over 300,000 equations. Overall, 140 Open Access documents are added daily to our database. All of this is possible due to our automation systems which are constantly working in the background to bring our customers the best information possible.

Picture1.gif

When completed, Orion will be able to create summaries/reports/reviews, on demand, in an autonomous way without any human intervention (only the title, headings and relevant keywords related to the summary/report requested will be necessary). This means that researchers and analysts will be able to get a "ready" to read/deliver report and access the latest advancements in the scientific world in a matter of minutes. "Orion" is being built to deliver such reports/summaries in a .docx (Microsoft word) document. "Orion" will be capable of screening thousands of papers in minutes (humanly impossible) and present the information in a comprehensive and accurate manner.

Visit Orion's repository here!

Below, there is a list of videos showing a basic version of Orion in action, selecting the top 20 best match papers for a variety of topics (from our database which by August 2021 was composed of 30,000 open access documents) and creating a .doc file in a fully autonomous way. The generated .doc file contains all the necessary information to get access to the papers. This includes paper title, authors, date of publication, journal, digital object identifier as well as the first image that appears in the corresponding paper.