<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Subhash Chandra on Medium]]></title>
        <description><![CDATA[Stories by Subhash Chandra on Medium]]></description>
        <link>https://medium.com/@subhashc.iisc?source=rss-d73e4753c51b------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*LWi2PPj5BXKW-grn</url>
            <title>Stories by Subhash Chandra on Medium</title>
            <link>https://medium.com/@subhashc.iisc?source=rss-d73e4753c51b------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 07:55:36 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@subhashc.iisc/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Lost in the Fog: My Journey into Radar-Based SLAM]]></title>
            <link>https://medium.com/@subhashc.iisc/lost-in-the-fog-my-journey-into-radar-based-slam-c4232ff42fd3?source=rss-d73e4753c51b------2</link>
            <guid isPermaLink="false">https://medium.com/p/c4232ff42fd3</guid>
            <dc:creator><![CDATA[Subhash Chandra]]></dc:creator>
            <pubDate>Tue, 25 Mar 2025 04:02:38 GMT</pubDate>
            <atom:updated>2025-03-25T04:02:38.216Z</atom:updated>
            <content:encoded><![CDATA[<p>By Subhash Chandra · Graduate Researcher · University of Oklahoma</p><p>It all started with a question that kept nudging me every time I saw an autonomous car commercial or a drone soaring across the sky: <em>What happens when the world isn’t clear?</em><br> When it’s foggy, rainy, snowy — or just plain dark?</p><p>As a researcher in computer science and robotics, I knew that most robots rely on either cameras (Visual SLAM) or lasers (LiDAR SLAM) to find their way — a technique known as <strong>Simultaneous Localization and Mapping</strong>, or simply, SLAM. But here’s the catch: cameras need light, and lasers can be thrown off by heavy fog or snow.</p><p>So I asked myself: <em>What about radar?</em> It’s the same tech airplanes use to land in storms and ships use in the dead of night. Could radar help robots navigate when the world around them disappears?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*l_5Jeqj_r6KFlXDO" /><figcaption>How SLAM will work in adverse weather</figcaption></figure><h3>The Summer I Went Radar Deep</h3><p>This past summer, I dove headfirst into the world of <strong>radar-based SLAM</strong>. I started reading every paper I could find. My room looked like a conspiracy wall of diagrams and radar returns. But three papers stood out and really shaped my journey:</p><ul><li><strong>RadarSLAM: Large-Scale SLAM in All Weathers</strong><br> This one was eye-opening. It showed how radar can see through snowstorms and heavy rain — situations where cameras just give up. It felt like radar was the quiet hero nobody talked about.</li><li><strong>ORORA: Outlier-Robust Radar Odometry</strong><br> Radar is noisy — that’s the trade-off. But this paper tackled that challenge head-on, using a method to separate rotational and translational movement in radar scans. It felt like cleaning a foggy mirror and suddenly seeing your reflection clearly.</li><li><strong>Scan Context</strong><br> Even though it was originally about 3D LiDAR, this paper introduced a clever way to recognize places the robot had already visited — even if it saw them from a completely different angle. That concept, I realized, could be borrowed for radar too.</li></ul><h3>From Paper to Practice: NavTech Radar in My Hands</h3><p>Reading is one thing. But doing? That’s a whole different story.<br> So I got my hands dirty — quite literally — setting up a <strong>NavTech radar sensor</strong>. This wasn’t plug-and-play. I had to calibrate it, process the raw radar returns, and feed it into a SLAM system.</p><p>The first few times, my maps looked like abstract art. Noisy blobs everywhere. But slowly, things started making sense. I began to “see” through the radar, understanding how it perceived the world. Every glitch, every spike of noise — it was all part of the learning curve.</p><p>I realized that while radar is powerful, it needs a good listener — algorithms that can filter out the noise and focus on the truth underneath.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*pUkjMI2dguF8O6Ud" /><figcaption>Photo by <a href="https://unsplash.com/@oregonstateuni?utm_source=medium&amp;utm_medium=referral">The Oregon State University Collections and Archives</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h3>What I Learned (and Why It Matters)</h3><p>Radar is resilient. It doesn’t care about light, or fog, or snow. It just works.<br> But it’s messy — kind of like life.</p><p>Compared to traditional SLAM methods:</p><ul><li><strong>Visual SLAM</strong> is amazing in sunny, well-lit places, but blind in the dark.</li><li><strong>LiDAR SLAM</strong> is super accurate, but struggles in blank, featureless environments — and it’s expensive.</li><li><strong>Radar SLAM</strong>? It’s the rugged survivor. Maybe not the prettiest, but the one you want when things go sideways.</li></ul><p>The real magic, I think, lies in combining them. A camera, a LiDAR, and a radar — all feeding their view of the world into a smarter, more adaptable brain.</p><h3>What’s Next?</h3><p>I’m now dreaming of <strong>multi-sensor fusion</strong> — letting radar team up with visual and LiDAR inputs to create an all-weather, all-terrain navigation system. I’m also exploring <strong>graph-based SLAM</strong>, where each location is like a node in a mind map, connected by mathematical relationships.</p><p>And there’s still so much to improve — like better noise filtering, smarter data association, and even multiple radars working together.</p><h3>Final Thoughts</h3><p>This journey into radar-based SLAM taught me something important:<br> Robustness isn’t just about being the best when things are easy — it’s about showing up when everything is falling apart.</p><p>That’s what radar does.<br> And in some way, I think that’s what I want to do too — build systems that don’t just work in perfect conditions, but thrive in the real world, with all its chaos.</p><p><em>If this story resonated with you, or if you’re working on autonomous systems, robotics, or sensor fusion — let’s connect. Drop me a line at subhashc.iisc@gmail.com, or follow me here for more of my research adventures.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c4232ff42fd3" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>