NRFL National Robot Football League Robots make NFL
NRFL - National Robot Football League Robots make NFL increasingly good
TOC NFL Gets Smarter Business Model How the robot is conceived Public Involvement How NFL Robots look like Robots better than humans ? Features . . . and cyborgs ? Training Emotions and Health Games Rules
Scenario - NFL Gets Smarter - New technology for NFL American Football association in 2030 - Idea of providing better players - Presentation of the model according to: - Features - Training issues - Emotions and health - Game’s rules - Cultural effect - Business model
How robot is conceived 01 CONTROLLER - also known as the brain, which is run by a computer program. Normally the program is very detailed as the developers set its movements and actions 02 MECHANICAL PARTS - motors, gripper, pistons and gears that make the robot move, grab, turn and lift. These parts are powered by air, water or electricity 03 SENSORS - to tell the robot about its surroundings. Sensors allow the robot to determine sizes, shapes, space between objects, direction and other relations and properties of substances. They are even able to identify the amount of pressure necessary to apply to grab an item without crushing it.
THE CONCEPT
Features Dynamic features: Total features = 53 - Mental - 10 statistical, related to the physical - Movement - Running - Blockage - Kickage - Throwage which range from 1 to 99 (assessed - Catchage every year) - Passage dimension and parameters - 43 dynamic, classified in the previous 8 groups - - Each group has some sub-factors Tendency and style
Training - Basic configuration settings - The robot is trained with a knowledge base imposed by the developer team - - Game’s rules - Basic movements - Other essentials Daily training based on Machine Learning divided in 2 steps: - Training (every day by the dev. team) - Testing (only during matches) - One code but a lot of improvements - Robots are stronger, faster and smarter - Each team has its own development team so that they can be differentiated
Emotions - Robots similar to humans ? - How is it possible to perceive emotions similar to humans ? - Real-time topic and Sentiment analysis in conversations - Interface that detects the topic and sentiments of a user’s utterances from text-transcribed speech
Perceiving Emotions - Back End TOPIC DETECTION SUBSYSTEM - Implemented on the robot’s onboard SENTIMENT ANALYSIS SUBSYSTEM - computer - - Latent Semantic Indexing model of a Implemented on the robot’s onboard computer - Disambiguate the input words, obtain conversation corpus, followed by a their sentiment scores from search in the online database Senti. Word. Net and returns the Concept. Net 5, in order to generate a averaged sum of the scores as the topically relevant reply overall sentiment score http: //conceptnet. io/ - https: //sentiwordnet. isti. cnr. it/
Health - Robots are designed to take on repetitive tasks (extreme precision) - Why better robots than humans ? a. 70% drop in hospital acquired infections b. Transport of an incredible amount of medical necessaries c. Interaction with patient increased (e. g. transport) d. Stress reduction (e. g. Paro) e. Robot’s life expectancy (20 years more or less) - Robots able to locate fastly what the problem is and find a remedy - Extremely trained by the developer team in order to answer to any supply - Machine learning algorithms (always training) - Precise sensors in order to substitute pieces to the damaged robot
Games Rules - 1 match: 80’ (4 quarters + 3 breaks) - 20 robots per team (11+9) - Robot out of battery: recharge and enter in next quarter - Abilities’ improvements ONLY ALLOWED by training and playing the VR Game - Code check by external dev. team of coaches 2 hours before the match - Referees: basic rules implied by machine learning algorithms
Business Model - NFL 8% profit decline (sign of the times) - The term ‘sport’ is defined differently in NRFL - Create a new paradigm of how you see, feel and participate in a modern concept of sport - New interaction system - better people-to-robot connection
Revenue Streams - Streaming of the Matches - Advertisement into the interaction system + during the matches - Tickets - Merchandising - Investments - Franchising - Globalized e-sport model - Lower operating cost
Interaction with public - Be part of the team !! - Coins for boosting robots - VR console - Sponsored content - Battle of fan bases - Simple fan or player concepts explained
“Football belongs to those who live it” - VR to be IN the game - Let’s play in millions ! - Huge team, huge love for this sport
NRFL - Dev. Team BRUNO MARAFINI YUE SONG LORENZO TAIT ANDRIY TAVRIN NHATDM ABHISHEK (critical mind)
Concept. Net Framework
Senti. Word. Net Resource
- Slides: 18