Jekyll2021-05-11T03:06:47+00:00http://jderobot.github.io/2019-tfg-natalia-monforte/feed.xmlMy TFG’s blogProgramming Robot IntelligenceNatalia Monforte RodríguezWeeks 39 & 402020-07-06T00:00:00+00:002020-07-06T00:00:00+00:00http://jderobot.github.io/2019-tfg-natalia-monforte/your%20category/week-39-40Tony StarkIntegration of Auxiliary Physics into ProductionWeeks 36, 37 & 382020-06-08T00:00:00+00:002020-06-08T00:00:00+00:00http://jderobot.github.io/2019-tfg-natalia-monforte/your%20category/week-36-37-38<p>This week’s tasks have been divided on two fronts:</p>
<ul>
<li>Creating a new exercise for the drone. This is a maze in which two versions have been implemented. The first one does not include signalling (the instructions used will indicate the distance to be covered). The second includes signs on the walls that the drone will have to interpret.</li>
<li>Deepening on the compression of the a-frame loop and the physics loop: The loop is unique. In each iteration of the a-frame rendering loop, CANNON code is executed in which positions and speeds are updated. Therefore, the problem can be solved by creating an auxiliary component that counts the number of ticks that are executed.</li>
</ul>Tony StarkNew Drone's exercises and Deepen the physics engine of A-FrameWeek 352020-06-01T00:00:00+00:002020-06-01T00:00:00+00:00http://jderobot.github.io/2019-tfg-natalia-monforte/your%20category/week-35<p>During this week I have been working on the integration of the new exercises:</p>
<ul>
<li>3D Labyrinth</li>
<li>Competitive football</li>
<li>Follows Sophisticated Lines</li>
</ul>
<p>I have also continued researching on possible A-Frame callbacks to get the rendering ticks of the A-Frame Loop counted.</p>Tony StarkNew exercisesWeek 342020-05-25T00:00:00+00:002020-05-25T00:00:00+00:00http://jderobot.github.io/2019-tfg-natalia-monforte/your%20category/week-34<p>During this week I continued with the problems of multi-robot physics. To do this, the following changes were made to the code:</p>
<ul>
<li>Calculation of the resulting velocity with a square root. In this way, we remove the dependence of sine and cosine.</li>
<li>Creation of an object for each robot that includes all the variables used in the auxiliary motor.</li>
</ul>Tony StarkMulti-object madness (Part II)Week 332020-05-18T00:00:00+00:002020-05-18T00:00:00+00:00http://jderobot.github.io/2019-tfg-natalia-monforte/your%20category/week-33<p>During this week I solved the problem of “multi-object madness”. I found two problems in the code:</p>
<ul>
<li>Division by 0 when the robot reached certain rotations.</li>
<li>Understand how CANNON acts on the position and speed update tick. This tick updates position and speed for all models in the world.</li>
</ul>Tony StarkMulti-object madness (Part I)Week 322020-05-11T00:00:00+00:002020-05-11T00:00:00+00:00http://jderobot.github.io/2019-tfg-natalia-monforte/your%20category/week-32<p>This week I advanced the project in the following aspects:</p>
<ul>
<li>Test of collisions with balls of different masses and restitutions.</li>
<li>Update of the A-Frame version.</li>
<li>Study of the code of websim. Especially, the implementation of simcore.</li>
</ul>Tony StarkMore collision's tests and SIMCOREWeek 312020-05-04T00:00:00+00:002020-05-04T00:00:00+00:00http://jderobot.github.io/2019-tfg-natalia-monforte/your%20category/week-31<p>During the week I have performed the following tasks:</p>
<ul>
<li>Associate a callback to object collisions in A-Frame. A callback is generated for each object and different properties of the collision can be studied.</li>
<li>Collision tests. Varying masses, restitutions, scenarios, etc.</li>
</ul>Tony StarkCollision's events and testsWeek 302020-04-27T00:00:00+00:002020-04-27T00:00:00+00:00http://jderobot.github.io/2019-tfg-natalia-monforte/your%20category/week-30<p>This week I made the following progress:</p>
<ul>
<li>Study of the AMMO physics engine. It seems that A-Frame will end up leaving aside CANNON and focusing more on AMMO, since it has a better maintenance of the code. Both engines are very similar, even AMMO is based on CANNON.</li>
<li>Begin to include the physics engine itself in the football exercise.</li>
<li>Test the CANNON physics in A-Frame, this time without touching up autonomous forces.</li>
</ul>
<p>Translated with www.DeepL.com/Translator (free version)</p>Tony StarkAMMO.js, friction and realistic physics on the football sceneWeek 292020-04-20T00:00:00+00:002020-04-20T00:00:00+00:00http://jderobot.github.io/2019-tfg-natalia-monforte/your%20category/week-29<p>This week I studied friction and collisions in a-frame.</p>
<ul>
<li>
<p>FRICTION:
Differentiation between static and dynamic friction Testing with our own engine. The friction attribute allows to model the friction in an a-frame world.</p>
</li>
<li>
<p>COLLISIONS:
Differentiation between static and dynamic collision. Carrying out tests without the engine itself. The restitution attribute allows the simulation of these two types of collision. Elastic collisions transfer 100% of the energy to the colliding body and inelastic collisions maintain 100% of the energy in the colliding body.</p>
</li>
</ul>Tony StarkFriction and collisionsWeek 282020-04-11T00:00:00+00:002020-04-11T00:00:00+00:00http://jderobot.github.io/2019-tfg-natalia-monforte/your%20category/week-28<p>This week I worked on the study of movement on the ramps of a stage. The results were as follows:</p>
<ul>
<li>In free fall, without an autonomous force that prevents the robot from falling down the ramp, the acceleration of the robot’s descent does not change if we change the mass of the robot.</li>
<li>If the friction of the world is increased, the robot advances more slowly and if it is dropped down a ramp without an autonomous force, it falls more slowly than when the world has a lower friction.</li>
</ul>
<p>Here is a video showing the test:</p>
<p><a href="https://www.youtube.com/watch?v=oxpPLnawyyY"><img src="http://img.youtube.com/vi/oxpPLnawyyY/0.jpg" alt="Prueba 1 fricción" /></a></p>Tony StarkFriction tests on ramps