* **Reading: <a href="/content/topics/homeworks/hw_readings_119/200_hw_119_brookspaper/">Read Brooks Paper</a>** *Reminder: Readings are your responsibility. You will be expected to come to class prepared, having read the material, and ready to participate in the discussion* ## Logistics * Overall impressions of wall follower submisions * Many algorithms looked similar. This is likely because of google, chatgpt, or student cooperation * And if that's as far as it went, that's perfect * You found that algorithms that work in sim dont work as is in real life. This is a well known fact. * The sim is never going to reproduce the quirks of a TB3 * Try to modularize more * Write functions, methods, classes or nodes to represent separable algorithms * Especially Nodes * Lets slook at the Lidar Scan Message a little closer. <slide_break></slide_break> ## Localization * I will be using these [Localization Slides](/content/topics/documents/f1slides.pdf) and information from [this very useful and in depth paper](https://www.cs.cmu.edu/afs/cs/project/jair/pub/volume11/fox99a-html/jair-localize.html) * Probabilistic algorithm * See figure in PowerPoint * Belief, Sense, Update ... etc. * But how is the belief represented, and how is the update done efficiently <slide_break></slide_break> ### AMCL Particles Display <img src="/content/topics/images/montecarlo.jpg" class="img-thumbnail d-block mx-auto" style="height:300px;"/> <slide_break></slide_break> ### Monte Carlo * **AMCL**: Advanced Monte Carlo Localization** (I've also seen "Augmented" and "Adaptive". Go figure.) * Why Monte Carlo? That's where the Casinos are! * Algorithms that incorporate random guesses when a direct solution is hard or not feasible * Diagram on the board how you would calculate Pi using a Monte Carlo Algorithm * Note for `PI` it's a very inefficient way to get an accurate result * But it illustrates the idea of Monte Carlo estimation <slide_break></slide_break> ### Particle Filter * Lets watch a short [Video About Particle FIlters](https://youtu.be/YBeVDxTHiYMz) * Markov localization means that the new state is dependent only on the previous state (and not the history) and that the probability distributions are * Markov localization = state estimation from sensor data * Instead of "solving" the equation for all data and all points * Use a Monte-Carlo technique * Generate a random collection of candidate locations * Compute the motion * Adjust the probability of each particle <slide_break></slide_break> ### Further references * [Video about Monte Carlo](https://youtu.be/ELetCV_wX_c) * You might be interested in these [Advanced Localization Slides](/content/topics/documents/localization.pdf). * And another useful link: [Where am I](https://medium.com/@NickHortovanyi/where-am-i-6cc0f6608c4c) * And another great explanation: [Hector Mapping](https://youtu.be/Q4qM-Uzj1SI) <slide_break></slide_break> <h2 class="shadow p-3 bg-white rounded">Thank you. Questions?<img class="img-fluid w-100" src="https://picsum.photos/800/100.jpg" /> <small> (random Image from picsum.photos)</small></h2>