<?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[Ati Motors - Medium]]></title>
        <description><![CDATA[The Ati Blog - Medium]]></description>
        <link>https://medium.com/ati-motors?source=rss----de1ce046997d---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>Ati Motors - Medium</title>
            <link>https://medium.com/ati-motors?source=rss----de1ce046997d---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Mon, 16 Nov 2026 20:51:23 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/ati-motors" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[DSLR: Dynamic to Static LiDAR Scan Reconstruction Using Adversarially Trained Autoencoder]]></title>
            <link>https://medium.com/ati-motors/dslr-dynamic-to-static-lidar-scan-reconstruction-using-adversarially-trained-autoencoder-feb2c53fdd1c?source=rss----de1ce046997d---4</link>
            <guid isPermaLink="false">https://medium.com/p/feb2c53fdd1c</guid>
            <category><![CDATA[aml]]></category>
            <category><![CDATA[autonomous-vehicles]]></category>
            <category><![CDATA[lidar]]></category>
            <dc:creator><![CDATA[Saurabh Chandra]]></dc:creator>
            <pubDate>Tue, 06 Jul 2021 10:32:45 GMT</pubDate>
            <atom:updated>2021-07-06T10:32:45.070Z</atom:updated>
            <content:encoded><![CDATA[<p>Ati Motors has a research tie up with IISc, Bangalore. Recently we published a paper together with them on improving LIDAR maps created in dynamic environments. Ati’s autonomous technology works together with humans and other equipment or vehicles in customer locations so this becomes an important aspect for us. The paper was accepted for publication in a top ranked conference: Association for the Advancement of Artificial Intelligence (AAAI 2021).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*huGxxpMmOfd4wDTX5H7GVw.png" /><figcaption>LiDAR view from the Sherpa at a factory</figcaption></figure><p>Specifically, we have shown that the maps generated in dynamic environments are more corrupt than a map generated in static environments. This work attempts to solve this problem by transforming a given dynamic LiDAR scan to a corresponding static LiDAR scan.</p><p>We pose the problem of Dynamic to Static Translation (DST) for LiDAR scans. We seek to learn the mapping from dynamic scans to the corresponding static scan to minimize the reconstruction loss. We focus on 2 two significant goals: accurately reconstructing the static structures like walls or poles and inpainting the occluded regions with a static background. Works performing DST for images and object point clouds [1–5], when adapted to our problem, produce suboptimal reconstructions for LiDAR data. LiDAR reconstruction works [6] come closest to our work to generate LiDAR scans using generative modeling. However, the reconstruction obtained using this work is not good enough for SLAM.</p><p>To address these shortcomings, we propose a Dynamic to Static LiDAR Reconstruction (DSLR) algorithm. It uses an autoencoder that is adversarially trained using a discriminator to reconstruct corresponding static frames from dynamic frames. Unlike existing DST-based methods, DSLR doesn’t require specific segmentation annotation to identify dynamic points. Additionally, if segmentation information is available, we extend DSLR to DSLR-Seg to further improve the reconstruction quality.</p><p>We train DSLR using a dataset that consists of corresponding dynamic and static scans of the scene. However, these pairs are hard to obtain in a simulated environment and even harder to obtain in the real world. To get around this, we propose two solutions: 1) Using Unsupervised Domain Adaptation, we propose DSLR-UDA for transfer to real-world data and experimentally show that this performs well in real-world settings. 2) A new paired-dataset generation algorithm generates appropriate dynamic and static pairs, making DSLR the first DST-based model to train on a real-world dataset.</p><p>Experiments on simulated and real-world datasets show that DSLR gives at least a 4x improvement over adapted baselines on the problem of DST for LiDAR. We also show that DSLR, unlike the existing baselines, is a practically viable model with its reconstruction quality within the tolerable limits for autonomous navigation tasks like SLAM in dynamic environments.</p><p>Please refer to our <a href="https://arxiv.org/abs/2105.12774">paper</a>, project pages (<a href="https://dslrproject.github.io/dslr/">link1</a> and <a href="https://sabyasachis.github.io/publication/dslr/">link2</a>), open-sourced <a href="https://github.com/dslrproject/dslr">code</a>, and <a href="https://github.com/dslrproject/dslr/tree/master/Data">datasets</a>.</p><p>References:</p><p>[1] Bescos et al. 2019. “Empty Cities: Image Inpainting for a Dynamic-Object-Invariant Space”. ICRA 2019.</p><p>[2] Chen et al. 2020. “Unpaired Point Cloud Completion on real scans using adversarial training”. ICLR 2020.</p><p>[3] Groueix et al. 2018. “AtlasNet: A Papier-Mâché Approach to Learning 3D Surface Generation”. CVPR 2018.</p><p>[4] Wu et al. 2020 “Multimodal Shape Completion via Conditional Generative Adversarial Networks”. ECCV 2020.</p><p>[5] Achlioptas et al. “Learning Representations and Generative Models for 3D Point Clouds”. ICML 2018.</p><p>[6] Caccia et al. 2019. “Deep Generative Modeling of LiDAR Data”. IROS 2019.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=feb2c53fdd1c" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ati-motors/dslr-dynamic-to-static-lidar-scan-reconstruction-using-adversarially-trained-autoencoder-feb2c53fdd1c">DSLR: Dynamic to Static LiDAR Scan Reconstruction Using Adversarially Trained Autoencoder</a> was originally published in <a href="https://medium.com/ati-motors">Ati Motors</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Sherpa Tug: Second Gen]]></title>
            <link>https://medium.com/ati-motors/sherpa-tug-second-gen-17592e6c9c80?source=rss----de1ce046997d---4</link>
            <guid isPermaLink="false">https://medium.com/p/17592e6c9c80</guid>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[autonomous-vehicles]]></category>
            <category><![CDATA[industrial-robotics]]></category>
            <dc:creator><![CDATA[Saurabh Chandra]]></dc:creator>
            <pubDate>Fri, 01 Jan 2021 12:52:39 GMT</pubDate>
            <atom:updated>2021-01-01T12:52:39.407Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*Hy3m-jFJRj-g8oyPTyc6nA.gif" /><figcaption>Sherpa Tug 2nd Gen</figcaption></figure><p>December has been an exciting month at Ati. We shipped our first second gen tug to a customer. We also shipped our first unit outside India, to our partner in Doha, Qatar for demoes to customers there. We booked orders, pilots at large OEMs and also started our CE certification process.</p><p>Earlier we utilised the pandemic for product improvements. Our deployment in a COVID hospital for contactless patient care made us add more remote deployment and monitoring features. Overall, the Second Gen platform has:</p><p>- completely new compute architecture that packs in the same computation at 1/10 the energy.<br>- Remote drive for support and deployment assistance. Works even over 4G.<br>- New compact battery design with 21700 cells. Can do single handed swaps now.<br>- Compact charge and swap stations that consume less space<br>- Move to a hub motor based drive system reducing moving parts and simplifying maintenance.<br>- Complete refresh of underlying electronics to be easy to service by field engineers<br>- New deployment tooling that allows field personnel to edit and make maps visually.<br>- New lidar mount for better protection against accidental damages and improved bidirectional movement<br>- New dynamic path planning for superior obstacle avoidance</p><p>It still has the same awesome indoor and outdoor capability not just algorithmically but as a physical vehicle with suspension even with the compact form factor we have achieved.</p><p>This year, we will add a 2 tonne variant to our catalog and also start work on a pallet mover. 2021, here we come.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=17592e6c9c80" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ati-motors/sherpa-tug-second-gen-17592e6c9c80">Sherpa Tug: Second Gen</a> was originally published in <a href="https://medium.com/ati-motors">Ati Motors</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Sherpa Mini]]></title>
            <link>https://medium.com/ati-motors/sherpa-mini-a65672f33cab?source=rss----de1ce046997d---4</link>
            <guid isPermaLink="false">https://medium.com/p/a65672f33cab</guid>
            <category><![CDATA[covid19bot]]></category>
            <category><![CDATA[bangalore-startups]]></category>
            <category><![CDATA[autonomous-vehicles]]></category>
            <dc:creator><![CDATA[Saurabh Chandra]]></dc:creator>
            <pubDate>Thu, 04 Jun 2020 19:24:46 GMT</pubDate>
            <atom:updated>2020-06-04T19:24:45.920Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FnqMNbk1p-e_riR__onOMw.png" /><figcaption>Sherpa Mini</figcaption></figure><p>Been a long hiatus on this blog and short business updates on LinkedIn seemed like a better channel for some time. We are back though. For those who follow us only here, a recap: we went commercial and closed first sale late last year. Deployed in a factory in Chennai in Jan 2020 where the Sherpa pulled nearly a ton of trolleys over a 400m round trip autonomously.</p><p>Our focus on making a vehicle capable of rugged conditions paid off quickly. The deployment environment was tough with poor lighting, no lane markers and very rough flooring. No typical industrial robotics kind of AGV would work in that place at all.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FEcow6ex6D6M%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DEcow6ex6D6M&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FEcow6ex6D6M%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/91dc522b139812289d1734fbb8b5a616/href">https://medium.com/media/91dc522b139812289d1734fbb8b5a616/href</a></iframe><p>A robust inbound customer interest meant that we had several orders waiting to get finalized once the new financial year started in April. But then the pandemic happened. The pandemic will result in a positive push for automation in general and autonomy in particular, given the social distancing requirements and worker shortages. However, in the short term the economy has a huge demand shock that will slow everything. So, we did what every startup wants to do right now — align supply to where the demand is.</p><p>We had been working on a new model for our autonomous vehicle — the Sherpa Mini. The requirement for this version has been echoed by many prospects over the last 12 months — for small payloads, narrow aisle or to carry sensors (like a thermal camera, 360 Camera, Gas Sensor, RFID reader etc). We were focused though on first getting a model to production before making another one.</p><p>We started on this model just before the pandemic lockdown started. With the pandemic, we were able to see an immediate use case of this form factor in hospitals as a medicine/food mover or disinfectant. Or when integrated with a thermal camera, a COVID-19 screener in places like airports.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FogwLFPaC6Rc%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DogwLFPaC6Rc&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FogwLFPaC6Rc%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/6c989fbc001c9c46c5f6f5cc36b86a84/href">https://medium.com/media/6c989fbc001c9c46c5f6f5cc36b86a84/href</a></iframe><p>The idea is to make it versatile. The top plate will allow multiple attachments like trays or connecting a disinfectant or sensors. We have power and connectivity available for attachments (USB, RS485, 12v DC)</p><p>There are some clever ideas in this model:</p><ul><li>the 3D lidar placement in the center gives us a 360º view. The top plate is placed such that all beams escape. We now support reversing also!</li><li>the rear single wheel is both drive and steering. Making it simpler and cost effective.</li><li>battery swap to allow 24x7 use</li><li>TeleOps over 4G (for remote support — this was essential since we want to avoid entering COVID wards). We already had OTA in the stack.</li></ul><p>The autonomy stack is exactly the same as the larger vehicle. The team has done a great engineering job to fit the same electronics in this form factor and this was realized during the lockdown!</p><p>Ati has tied up with Wipro for a go-to-market in the healthcare space. The hospital version has telemedicine features, equipment to take vitals and space to carry food or medicine to patients. A trial has already started last week at a designated COVID-19 hospital in Bangalore.</p><p>Wish us luck!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a65672f33cab" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ati-motors/sherpa-mini-a65672f33cab">Sherpa Mini</a> was originally published in <a href="https://medium.com/ati-motors">Ati Motors</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Sherpa in Motion]]></title>
            <link>https://medium.com/ati-motors/sherpa-in-motion-de56fbcce845?source=rss----de1ce046997d---4</link>
            <guid isPermaLink="false">https://medium.com/p/de56fbcce845</guid>
            <category><![CDATA[bangalore-startups]]></category>
            <category><![CDATA[autonomous-vehicles]]></category>
            <dc:creator><![CDATA[Saurabh Chandra]]></dc:creator>
            <pubDate>Thu, 01 Aug 2019 11:08:11 GMT</pubDate>
            <atom:updated>2019-08-01T11:17:49.068Z</atom:updated>
            <content:encoded><![CDATA[<p>The bits have become atoms and they move.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FLnhaZ5FdlLs%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DLnhaZ5FdlLs&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FLnhaZ5FdlLs%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/730bf6ba4789dce0ba6d60462f7aeb60/href">https://medium.com/media/730bf6ba4789dce0ba6d60462f7aeb60/href</a></iframe><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*W5sOz6OKpiOOVbtMzFS4xQ.jpeg" /></figure><p>We have done many customer trials with the previous prototype. Now going into commercial deployments with this production version. Best part — huge incoming interest from customers!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=de56fbcce845" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ati-motors/sherpa-in-motion-de56fbcce845">Sherpa in Motion</a> was originally published in <a href="https://medium.com/ati-motors">Ati Motors</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Sherpa upgrades]]></title>
            <link>https://medium.com/ati-motors/sherpa-upgrades-57a276799028?source=rss----de1ce046997d---4</link>
            <guid isPermaLink="false">https://medium.com/p/57a276799028</guid>
            <category><![CDATA[india]]></category>
            <category><![CDATA[autonomous-vehicles]]></category>
            <category><![CDATA[startup]]></category>
            <dc:creator><![CDATA[Saurabh Chandra]]></dc:creator>
            <pubDate>Mon, 18 Mar 2019 05:30:22 GMT</pubDate>
            <atom:updated>2019-03-18T05:30:22.492Z</atom:updated>
            <content:encoded><![CDATA[<p>We did a minor update of our 1.5 version with some practical features like a split design and hood with a hinge for easy maintenance. That got us to 1.6.</p><p>But the big change is coming in the next version — 1.7 We are getting rid of the bump in the rear. That makes it a full flat bed on the back giving us much more flexibility in adding loading options. There is a lot to talk about this new design but in due time.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DGyUX5UmS1ptnum8gcyI0g.png" /></figure><p>If you follow us elsewhere, you may have seen us getting a lot of social media love in the last few months. It has led to more than encouragement — the number of genuine customer and partnership enquiries is making us pay much more attention to the Indian market. The sweet spot arising from plummeting technology costs and affordability of customers in India is getting close.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=57a276799028" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ati-motors/sherpa-upgrades-57a276799028">Sherpa upgrades</a> was originally published in <a href="https://medium.com/ati-motors">Ati Motors</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[In-place Turn]]></title>
            <link>https://medium.com/ati-motors/in-place-turn-60eb1a473461?source=rss----de1ce046997d---4</link>
            <guid isPermaLink="false">https://medium.com/p/60eb1a473461</guid>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[autonomous-vehicles]]></category>
            <category><![CDATA[maneuverability]]></category>
            <dc:creator><![CDATA[Saurabh Chandra]]></dc:creator>
            <pubDate>Tue, 09 Oct 2018 18:28:53 GMT</pubDate>
            <atom:updated>2018-10-09T18:28:52.731Z</atom:updated>
            <content:encoded><![CDATA[<p>We have mentioned earlier how our unique design gives us better maneuverability, including the ability to turn in-place. See it in action.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FG-G5D1D09u0%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DG-G5D1D09u0&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FG-G5D1D09u0%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/90a5604027ab445a647e08dacaf8ade3/href">https://medium.com/media/90a5604027ab445a647e08dacaf8ade3/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=60eb1a473461" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ati-motors/in-place-turn-60eb1a473461">In-place Turn</a> was originally published in <a href="https://medium.com/ati-motors">Ati Motors</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Sherpa goes Go-Karting]]></title>
            <link>https://medium.com/ati-motors/sherpa-goes-go-karting-ce70b475fe80?source=rss----de1ce046997d---4</link>
            <guid isPermaLink="false">https://medium.com/p/ce70b475fe80</guid>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[bangalore]]></category>
            <category><![CDATA[autonomous-vehicles]]></category>
            <dc:creator><![CDATA[Saurabh Chandra]]></dc:creator>
            <pubDate>Fri, 28 Sep 2018 16:41:25 GMT</pubDate>
            <atom:updated>2018-09-28T16:41:24.741Z</atom:updated>
            <content:encoded><![CDATA[<iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FZrN6NV7gzXE%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DZrN6NV7gzXE&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FZrN6NV7gzXE%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/bbb80188a70c2dc4c4d739fc973ab0ee/href">https://medium.com/media/bbb80188a70c2dc4c4d739fc973ab0ee/href</a></iframe><p>We have had the commercial prototype of Sherpa 150 with us for a few weeks now and it underwent similar tests as the prototype zero. However, we wanted to do something more and do longer runs at higher speeds than we have been able to do at the IISc campus. A Go-Kart track seemed like an ideal environment to do that, especially considering that there is one close to our office.</p><p>The key objectives were to analyse the drifts that we see in our localisation estimates over longer distances (turn out to be pretty good — you can see the vehicle threading the cones from where it started at the end), see our control algorithm response at various speeds (was getting better at higher speeds!) and have some fun at the Go-Karting track (I can skid a go-kart while taking turns, yea!). We were able to achieve all objectives.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ce70b475fe80" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ati-motors/sherpa-goes-go-karting-ce70b475fe80">Sherpa goes Go-Karting</a> was originally published in <a href="https://medium.com/ati-motors">Ati Motors</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The killer app for autonomy in India]]></title>
            <link>https://medium.com/ati-motors/the-killer-app-for-autonomy-in-india-904314ba531?source=rss----de1ce046997d---4</link>
            <guid isPermaLink="false">https://medium.com/p/904314ba531</guid>
            <category><![CDATA[ada]]></category>
            <category><![CDATA[solveforindia]]></category>
            <category><![CDATA[autonomous-vehicles]]></category>
            <dc:creator><![CDATA[Saurabh Chandra]]></dc:creator>
            <pubDate>Mon, 10 Sep 2018 08:03:50 GMT</pubDate>
            <atom:updated>2018-09-10T08:03:49.694Z</atom:updated>
            <content:encoded><![CDATA[<p>tldr; adas for two wheeler.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*Zi7l14EbXWig5rmQJH1XQg.jpeg" /><figcaption>2-wheelers are commute vehicles in India (image: wikipedia)</figcaption></figure><p>The biggest reason for autonomous vehicles to exist is safety. And while level 4 and 5 autonomy are eagerly awaited and make the most news, level 2 (which are driving assistance systems) have been a reality for many years now in western markets. Autonomous Driving Assistance Systems (ADAS) are already multi-billion dollar industries. While we don’t see much of these in India due to cost considerations, it is just a matter of time when they come here. However, startups in the west are focused on 4-wheelers.</p><p>This June, the number of 2-wheelers sold in India was more than 4 times all the 4 wheelers put together (18+ lakh vs 4 point something). This is a space that no valley based startup is going to solve for India. It is a minuscule market in the western countries where 2-wheelers are recreational vehicles unlike in India (and many other emerging markets) where they are commuting vehicles. This is an interesting space where an Indian startup could create a global solution by focusing on a very local problem.</p><p>How does one think of a solution in this space — obviously cost is a concern (most vehicles are under $1000) and the form factor of the vehicle is very different too. The driving assistance technology should not become a distraction that worsens the problem. The answer would lie in the helmet.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NPeQwO0xtI3YrhmJuMeFDQ.jpeg" /><figcaption>Motorcycle Crash Helmet</figcaption></figure><p>The only way to give ADAS cues inside a helmet such that the rider doesn’t take her eyes off the road, is to use heads-up displays. One can imagine an integrated solution where sensors are integrated with a vehicle and communicating with the visuals on the helmet through a wireless technology like bluetooth or it could be a completely independent solution with sensors integrated on the helmet itself. The latter solution can be most interesting since this could make any 2-wheeler a smart one with ADAS. A startup doesn’t need to tie up with specific OEMs and be restricted to a few brands to offer this solution. It also increases the value for a rider since the solution can get shared or used across vehicles.</p><p>While Ati is focused on a B2B market, we keep thinking of interesting problems in the space and since this blog has many readers interested in the overall autonomous market, hopefully someone will pick on this line of thought to make this a reality. Getting the energy, compute, sensors together with the thermal constraints worked in the form factor of a helmet seems like an awesome problem to crack.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=904314ba531" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ati-motors/the-killer-app-for-autonomy-in-india-904314ba531">The killer app for autonomy in India</a> was originally published in <a href="https://medium.com/ati-motors">Ati Motors</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Uber for Factory]]></title>
            <link>https://medium.com/ati-motors/uber-for-factory-ebb5d88a37c3?source=rss----de1ce046997d---4</link>
            <guid isPermaLink="false">https://medium.com/p/ebb5d88a37c3</guid>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[bangalore]]></category>
            <category><![CDATA[autonomous-vehicles]]></category>
            <category><![CDATA[uber-for-x]]></category>
            <dc:creator><![CDATA[Saurabh Chandra]]></dc:creator>
            <pubDate>Wed, 22 Aug 2018 19:27:45 GMT</pubDate>
            <atom:updated>2018-08-22T19:27:45.209Z</atom:updated>
            <content:encoded><![CDATA[<p>We are on for a pilot soon!</p><p>The first one is in a manufacturing facility and we have come up with this clever (ahem!) metaphor of an uber for factory user flow. While the eventual usage for a vehicle such as ours will be API driven (some operational or business application will call APIs on the vehicle), pilots can easily benefit from an uber like workflow. Material typically has to be moved between Stations A, B, C … N. We give an app to be used on a phone or tablet that allows requesting the vehicle to go from point A to N. Vehicle comes over, notifies. User loads stuff and hits start trip. On the other side, user gets notified that vehicle is on the way with an ETA. Receive stuff, end trip. Done.</p><p>Here is a small test we did mocking such a workflow in our test environment.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F_jhTBxm8RBk%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D_jhTBxm8RBk&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2F_jhTBxm8RBk%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/07a3ab483eca928dfeef6039ccd1c7df/href">https://medium.com/media/07a3ab483eca928dfeef6039ccd1c7df/href</a></iframe><p>What we see is that the application layer on top of the vehicle can become pretty specific to a vertical. For example, the above workflow won’t work in an eCommerce warehouse (where incidentally our next pilot is). Thankfully, the software stack on the vehicle is nicely decoupled from the application on top and we can easily create the right interfaces needed.</p><p>One more thing we saw coming into these pilots is that a more industrial vehicle like design is more suited in these environments and most of the use cases were within a 150kg payload capability. That prompted our team to spin a new vehicle in record time with a much more compact footprint for this smaller payload. Here is the design render, but we should have the real vehicle in showcase next month. Watch this space!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xSxJhEDROU9vzwa-asA-rA.png" /><figcaption>Sherpa 150</figcaption></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ebb5d88a37c3" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ati-motors/uber-for-factory-ebb5d88a37c3">Uber for Factory</a> was originally published in <a href="https://medium.com/ati-motors">Ati Motors</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A longer demo reel]]></title>
            <link>https://medium.com/ati-motors/a-longer-demo-reel-8b64086db576?source=rss----de1ce046997d---4</link>
            <guid isPermaLink="false">https://medium.com/p/8b64086db576</guid>
            <category><![CDATA[autonomous-vehicles]]></category>
            <category><![CDATA[startup]]></category>
            <dc:creator><![CDATA[Saurabh Chandra]]></dc:creator>
            <pubDate>Fri, 29 Jun 2018 18:33:13 GMT</pubDate>
            <atom:updated>2018-06-29T18:33:12.802Z</atom:updated>
            <content:encoded><![CDATA[<iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FhBFGQ1F70gY%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DhBFGQ1F70gY&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FhBFGQ1F70gY%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/baab7e67944323df882a5b269d7f4b65/href">https://medium.com/media/baab7e67944323df882a5b269d7f4b65/href</a></iframe><p>The vehicle keeps doing new stuff, so updated our demo reel. Watch obstacle avoidance (Naveen — who stands as the obstacle is perfectly distracted on his phone), overtaking and a mischief maneuver where we manually perturb the trajectory to see if the algorithm is robust enough to rejoin. It is!</p><p>We are now actively working on the next prototype while the software stack continues to mature on prototype zero. Mechanical and electronics design are both complete. Next step is to manufacture the chassis and then integrate the electronics. Eager to get that going since this version will carry real cargo and be ready for some early pilots.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8b64086db576" width="1" height="1" alt=""><hr><p><a href="https://medium.com/ati-motors/a-longer-demo-reel-8b64086db576">A longer demo reel</a> was originally published in <a href="https://medium.com/ati-motors">Ati Motors</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>