<?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 Louis Khalayli on Medium]]></title>
        <description><![CDATA[Stories by Louis Khalayli on Medium]]></description>
        <link>https://medium.com/@louiskh?source=rss-7f1716cc1864------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*nP_MhqQ6qyLpHuz9knJwnQ.jpeg</url>
            <title>Stories by Louis Khalayli on Medium</title>
            <link>https://medium.com/@louiskh?source=rss-7f1716cc1864------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 13:19:10 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@louiskh/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[Deep Learning to Check a Robot’s Work]]></title>
            <link>https://medium.com/swlh/deep-learning-to-check-a-robots-work-299dbf7b4a6a?source=rss-7f1716cc1864------2</link>
            <guid isPermaLink="false">https://medium.com/p/299dbf7b4a6a</guid>
            <category><![CDATA[robotics]]></category>
            <category><![CDATA[manufacturing]]></category>
            <category><![CDATA[quality]]></category>
            <category><![CDATA[computer-vision]]></category>
            <category><![CDATA[automation]]></category>
            <dc:creator><![CDATA[Louis Khalayli]]></dc:creator>
            <pubDate>Tue, 03 Jan 2023 13:39:30 GMT</pubDate>
            <atom:updated>2023-05-21T10:59:30.377Z</atom:updated>
            <content:encoded><![CDATA[<h4>or a human for that matter.</h4><p>It may sound like a scene from a sci-fi movie, but it’s not. Robotics have been incredibly useful in manufacturing since at least 1960’s (for the automotive industry). As technology improved, applications of robots in manufacturing have and continue to grow fast; however, one of the issues in using them is their potential tendency to go “out-of-spec” — meaning, they sometimes don’t meet the quality standards. Although, breached quality standards by robots happen at a much lower rate than humans, it is still very important for many purposes to catch every defect before it makes its way to customers (purposes such as customer safety, brand protection, regulation, etc.)</p><blockquote>So, practically — how does one make sure the one defected part in a thousand parts is contained? How can one cost-effectively inspect every single part coming out of a robot at a rate faster than humanly possible?</blockquote><p>You guessed it. <strong>More automation</strong>. Automated Quality Inspection of every single part. It becomes even <em>more important</em> in robot-less operations, as defect rates multiply several folds when humans are involved.</p><p>In the article below, I explain — based on my prior experience automating processes and quality inspections using robots and cameras at some of the world’s largest factories — how two different techniques for automated quality inspection compare to each other, and why Deep Learning wins, well.. every single day.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*6TyeHXY7tD7AlHA3" /><figcaption>Photo by <a href="https://unsplash.com/@lennykuhne?utm_source=medium&amp;utm_medium=referral">Lenny Kuhne</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Deep learning has become a popular approach for solving problems in computer vision, and for good reason. Unlike traditional, rules-based approaches, deep learning allows computers to learn and make predictions based on data, without requiring explicit instructions. This has several benefits, including improved accuracy, the ability to handle complex and unstructured data, and the ability to adapt to changing circumstances. This is especially valuable in several fields, such as automated quality inspection, where the demands and requirements may change over time.</p><p>One of the key advantages of deep learning over rules-based approaches is its ability to improve accuracy. Traditional computer vision systems rely on hand-coded rules, which are often brittle and can break when faced with complex or unexpected data. In contrast, deep learning algorithms are able to learn from data, allowing them to make more accurate predictions. This is particularly useful for tasks like object recognition, where deep learning algorithms have been shown to outperform traditional approaches by a significant margin.</p><blockquote>“This makes them particularly well-suited for tasks such as identifying defects in complex manufactured goods, where the appearance and location of defects may vary widely.”</blockquote><p>Another benefit of deep learning is its ability to handle complex and unstructured data. Traditional computer vision systems rely on structured data, such as images with clear, well-defined objects. However, in many real-world scenarios, data can be messy and unstructured, making it difficult for traditional algorithms to make sense of it. Deep learning algorithms, on the other hand, are able to handle a wide variety of data, including images, audio, and text, allowing them to make predictions even in complex and unstructured environments. This makes them particularly well-suited for tasks such as identifying defects in complex manufactured goods, where the appearance and location of defects may vary widely.</p><p>Another advantage of deep learning is its ability to adapt to changing circumstances. Traditional computer vision systems are brittle, and can break if the conditions they are operating under change. For example, a system trained to recognize objects in a specific environment may not be able to handle changes in lighting, background, or other factors. In contrast, deep learning algorithms can be trained on a large and diverse dataset, allowing them to adapt to changing conditions and make accurate predictions even in novel situations.</p><p>In conclusion, deep learning offers several benefits over traditional, rules-based approaches for computer vision. Its ability to improve accuracy, handle complex and unstructured data, and adapt to changing circumstances make it a powerful tool for a wide range of applications, from object recognition to autonomous vehicles.</p><p><em>In the spirit of automation, the latter part of this article was written by ChatGPT, edited and reviewed by Louis Khalayli &amp; Daniel Manary.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=299dbf7b4a6a" width="1" height="1" alt=""><hr><p><a href="https://medium.com/swlh/deep-learning-to-check-a-robots-work-299dbf7b4a6a">Deep Learning to Check a Robot’s Work</a> was originally published in <a href="https://medium.com/swlh">The Startup</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Confessions of a Bike Rider]]></title>
            <link>https://medium.com/@louiskh/confessions-of-a-bike-rider-5868faec0a2f?source=rss-7f1716cc1864------2</link>
            <guid isPermaLink="false">https://medium.com/p/5868faec0a2f</guid>
            <category><![CDATA[fundraising]]></category>
            <category><![CDATA[sports]]></category>
            <category><![CDATA[australia]]></category>
            <category><![CDATA[cancer-research]]></category>
            <category><![CDATA[bikes]]></category>
            <dc:creator><![CDATA[Louis Khalayli]]></dc:creator>
            <pubDate>Thu, 08 Oct 2020 04:20:56 GMT</pubDate>
            <atom:updated>2020-10-09T04:13:44.307Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Preparing for a 200 kilometer cancer-research-funding charity ride</em></p><p>I felt like my heart could stop any minute.</p><p>The force on the pedal became exponentially harder as e-v-e-r-y second passed.</p><p>If I were to open up my muscles,</p><p>I could pour lactic acid that would flood the freeway and cause a traffic crisis..</p><p>Perth wasn’t even visible in the horizon;</p><p>Yes, we were that far.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*8nQRQHW-p0_nki4M" /><figcaption>Photo by <a href="https://unsplash.com/@manuelmeurisse?utm_source=medium&amp;utm_medium=referral">Manuel Meurisse</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h4>All I could picture in my mind was..</h4><p>that fresh <em>custard</em> filled chocolate glazed donut in the servo fridge that I didn’t eat when I had the chance during the break. (servo = gas service station, shortened to 2 syllables in Aussie accent)</p><p>It was starting to get cloudy, drizzly, colder every second..</p><p>Pedal softly to rest? Not an option. For the headwind was merciless. I had to keep up with the rest of the group.</p><p>Drafting was essential!</p><p>My core exhausted, my shoulders sore..</p><p>Heck, my legs were numb..</p><p>I tried to distract myself with something to pedal o-n-e m-o-r-e tiiiiime. I imagined myself hiding under the doona feeling warm. (Doona = Australian English for warm bed cover/quilt)</p><p><em>Surprisingly, that didn’t help.</em></p><p>Then I looked down and noticed the floor passing me at a fast pace,</p><p>every pebble, every twig, every gum nut.. (gumnut = droppings of Eucalyptus trees, very hard and capable of flipping you <em>and your bike</em> so that the next thing you’ll be looking at is the blue skies — very common in the Terra Australis)</p><p>And my love for speed was quenched! That worked! Look down and be happy about how fast I’m going!</p><p>But oh well, you gotta focus on the road, mate.</p><blockquote>“Your back tyre is getting flat, maate!”</blockquote><p>My teammate yelled as she passed me by..</p><p>My mind went “KA-CHING”!</p><p>That was it! That’s why it was getting harder. It wasn’t <em>me</em> getting tired — I found an outlet for <em>all</em> of this ordeal.</p><p>It was the extra resistance of the tyre.</p><p>I got that sudden energy to push.</p><p>After all..</p><p>It was <em>all</em> in my head.</p><p>That was the toughest 150 km I’ve ever ridden on a weekend in my life.</p><p><em>No pain will ever be similar to that of cancer patients.</em></p><p><em>The Ride to Conquer Cancer is an international event where bike riders raise a considerable amount of money for cancer research and participate in a 200 km bike ride over 2 days to raise awareness.</em></p><p><em>Donate to your local cancer research facility.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5868faec0a2f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Electric Vehicle Conversion Shopping List and an Expensive Lesson in Business]]></title>
            <link>https://medium.com/swlh/electric-vehicle-conversion-shopping-list-and-an-expensive-lesson-in-business-ed82c0fde4f8?source=rss-7f1716cc1864------2</link>
            <guid isPermaLink="false">https://medium.com/p/ed82c0fde4f8</guid>
            <category><![CDATA[business]]></category>
            <category><![CDATA[conversion]]></category>
            <category><![CDATA[trends]]></category>
            <category><![CDATA[mechatronics]]></category>
            <category><![CDATA[electric-car]]></category>
            <dc:creator><![CDATA[Louis Khalayli]]></dc:creator>
            <pubDate>Mon, 28 Sep 2020 02:22:44 GMT</pubDate>
            <atom:updated>2020-10-07T20:54:18.996Z</atom:updated>
            <content:encoded><![CDATA[<p>A few years ago, I dreamt up an idea — If electric cars are a trend, and many <a href="https://en.wikipedia.org/wiki/Phase-out_of_fossil_fuel_vehicles">governments are banning the sales of internal combustion engines in their countries by 2025–2040</a> as part of the phase out of fossil fuel, wouldn’t it be a waste to eventually destroy all cars on the road?</p><p>With a strong fire of passion to make the world a better place while succeeding in business, I talked to a close friend about it — and we immediately thought it would be a <em>great</em> idea to convert cars to electric! So we set off to try one.</p><p>Who would have thought? A couple of guys converting a car to electric in the most isolated city in the world that no one has ever heard of — Perth, the capital of Western Australia? Well, it turns out, we weren’t the first: there were 15 (maybe more now) other converted cars on the road by a previously-at-BMW German professor at the local university.</p><p>Side note: Perth, a hidden gem, is the <a href="https://www.youtube.com/watch?v=NOXSSaiFYdM">most beautiful city</a> I’ve ever visited or lived in, with pristine beaches, daily sightings of dolphins in the river, and daytime temperatures around 20–28 degrees C year round.</p><p>The next few paragraphs are more technical, so feel free to skip to the last section on the business lessons.</p><p><strong>Electric car conversion basics</strong></p><p>My friend had an old 1977 Toyota Stout — you may know its successor, the Hilux truck. It was a 40-year-old rusty piece of junk, which my friend best described as “you wouldn’t even pick it up from the side of the road if it had a “FREE” sign on it”. There was fuel leaking from the tank, rust holes everywhere, and a seat that released 30-something years accumulation of dust when you sat on it *cough*cough*</p><blockquote>LESSON 1: Choose a car in better condition. Restoring a classic vehicle is a LOT of work.</blockquote><p>My friend’s wife hated the car because of how the whole garage smelled like gasoline (remember the leaky fuel tank?) so we started off by removing the leaky fuel tank, rendering the car immovable at the start of the project. Then the car engine and transmission, and eventually all the fuel related parts (fuel carburettor, fuel lines, gas pedal…) The car was getting a lot cleaner without those fossil fuel burning engines. Here is a photo of the result.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KL7oZ3yRtpNRSixGyi1GVA.jpeg" /><figcaption>Victory Moment!</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*L_31d68UGq9TqUCeOD68nA.jpeg" /><figcaption>Engine and Tranny Out (Transmission — in Aussie accent everything is 2-syllables)</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zcKrZEd0QANA3REp6-c_KA.jpeg" /><figcaption>Empty shell — no fuel tank, engine, carburettor, ..</figcaption></figure><p>While I make it sound simple, do not underestimate the number of special tools you need to perform this job. My friend had an engine crane lying around from a prior business (this guy had tools to start a construction company and a little electric car conversion garage — perfect partner for this project). At some stage, some of the older nuts were so tight and rounded (specifically the transmission one), we had to weld a metal bar to them to untighten them. Other bolts had to be ground off. Not for the faint hearted.</p><blockquote>LESSON 2: Refer to lesson 1. Oh and, you need an engine crane, a welding machine, an angle grinder, and lots of protective equipment.</blockquote><p><strong>Techy Electric car design stuff</strong></p><p>The funniest thing I’ve heard is people comparing electric cars to an electric toothbrush. While it is true that at a very simplistic level, both are made of a battery, a bunch of copper cables, and an electric motor. But being in car manufacturing, I can tell you — cars are one of the most complex machines ever made by humans (to put it into perspective, the smart phone, with all its glory, is only a small subset of a car — namely, the infotainment system, that touchscreen that provides you navigation and audio). Electric cars are a LOT simpler and a LOT more energy efficient to drive than your best internal combustion engine car — but they are not a simple electric toothbrush (<a href="https://www.bbc.com/news/business-50004184">and definitely not a Dyson vacuum machine either</a>).</p><blockquote>Side notes:</blockquote><blockquote>If a company does not know their “why”, the company would struggle to achieve its goals, as per Simon Sinek’s “Start With Why: How Great Leaders Inspire Everyone to Take Action”. Highly recommended read (<a href="https://www.amazon.com/Start-with-Why-Simon-Sinek-audiobook/dp/B074VF6ZLM">audiobook</a> is even better).</blockquote><blockquote>The worst electric car is usually about <a href="https://fueleconomy.gov/feg/evtech.shtml">77% percent efficient</a> (battery to wheels). The best internal combustion car is <a href="https://www.greencarreports.com/news/1091436_toyota-gasoline-engine-achieves-thermal-efficiency-of-38-percent">at about 40% efficient</a> (the rest is lost as heat). Your best happy medium is the hybrid.</blockquote><p>Even though I’m an electrical engineer with a master’s degree in mechatronics, one of the very early lessons I learnt in my career is “you don’t know what you don’t know” so it’s important to go through the learning phase. Two great intros to electric and hybrid car design (and their safety procedures) are in this <a href="https://www.amazon.ca/Build-Your-Own-Electric-Vehicle/dp/0071543732/ref=pd_lpo_14_t_1/144-1299338-1692306?_encoding=UTF8&amp;pd_rd_i=0071543732&amp;pd_rd_r=a220fc0c-7601-42b5-957a-bf8d5d9095e8&amp;pd_rd_w=S7GuH&amp;pd_rd_wg=2WWc5&amp;pf_rd_p=256a14b6-93bc-4bcd-9f68-aea60d2878b9&amp;pf_rd_r=R419296HYNHWM9XY17J7&amp;psc=1&amp;refRID=R419296HYNHWM9XY17J7">book</a> by Seth Leitman and Bob Brant, as well as <a href="https://mtasant.com.au/training/hybrid-and-battery-electric-vehicle-5-day-course">this course</a> by the Motor Trade Association of South Australia (at the time, it was the only course of its kind in South East Asia and Australia/New Zealand). I drove 2650 km across the Nullarbor to Adelaide to attend this course.</p><blockquote><em>Side lessons:</em></blockquote><blockquote>Do not stop at Kalgoorlie, Western Australia</blockquote><blockquote>Make sure you stop at Esperance, though, and Balladonia, where the <a href="https://www.abc.net.au/news/2020-05-31/nasa-skylab-fell-to-earth-esperance-retrofocus/12282468">NASA’s Skylab “met its demise” in 1979, after which NASA got fined $400 for littering by the Shire of Esperance</a></blockquote><blockquote>My 2006 Lexus IS 250 at 110 km/hr on highways could do close to 1000 kms on only <em>one full tank</em> of fuel (ironic, the presence of this note is.. in an article advocating for electric cars)</blockquote><p><strong>If you are converting an electric car, here is your (condensed) shopping list that’ll set you back some A$ 20k to A$ 35k (lead acid to lithium batteries)*:</strong></p><p>*Hint: don’t go to Woolies, Coles, or — if you’re in North America — Walmart</p><p>**You could do it for a lot cheaper using poor quality components</p><p>***Does not include many special tools</p><ul><li><em>An electric motor (preferably a 3-phase AC induction motor)</em></li></ul><p>We used a Solectria AC-55 we found at someone’s garage, pulled out from an old Ford EV</p><ul><li><em>An inverter (often called motor controller)</em></li></ul><p>We used an Azure Dynamics DMOC 445, compatible with the AC-55</p><blockquote>Big lesson here: this inverter was made by a Canadian company that had closed down years ago. Its second hand products are of good quality and still available on the market but they are no longer supported so you are on your own (these are very complex pieces of equipment, so without manuals, electrical drawings, software to connect to them, it is a <strong>nightmare</strong> to get to work — so they are cheap for a reason)</blockquote><ul><li><em>Batteries</em> (Ohhh mama…! The prices are insane)</li></ul><p>We used 12V lead acid (feeding about 336V to the inverter)</p><ul><li><em>A battery management system (BMS)</em></li></ul><p>Preferrably used for Lithium batteries</p><ul><li><em>High power rated (orange) cables and crimps (and a special hydraulic crimper)</em></li></ul><p>We got some from the local electrician shops</p><ul><li><em>Grounding cables</em></li><li><em>A new driveshaft to fit your motor</em></li></ul><p>This depends on the dimensions and design — we decided to modify our existing driveshaft, at a local heavy machines mechanic</p><ul><li><em>Circuit breakers, fuses, and switches</em></li><li><em>DC-DC converter (to charge your 12V battery from your HV battery)</em></li></ul><p>Doesn’t matter which one you use, as long as it’s compatible with your voltage level at your HV and LV (336V to 14V in our case)</p><ul><li><em>A multimeter (to measure voltages, etc.)</em></li><li><em>A 3-phase electrical AC compressor and inverter</em></li></ul><p>Normal car air compressors don’t work in electric cars — you need one that feeds from the battery — we skipped this</p><ul><li><em>Lots of jigs to fit things around</em></li><li><em>A charger board (to charge your HV battery from your 240V supply)</em></li><li><em>Ground fault detection boards</em></li></ul><p>Grounding is very important in electric cars, and is a very serious safety measure</p><ul><li><em>An accelerator pedal with a potentiometer</em></li><li><em>Lots of switches, wiring for electronics, etc….</em></li></ul><p>All of the above must be compatible with each other, and provide the right amount of torque to keep the car moving, so there is — as you would imagine — an intense design exercise, trying to match puzzle pieces together, without causing a short circuit that will burn the house down (I’m rather sure the insurance company wouldn’t cover electric car conversions as an acceptable cause).</p><blockquote>LESSON 3: Do NOT perform an electric car conversion in Western Australia. There are just no parts lying around for electric car conversions — we had to order most things from the US, China, East Coast of Australia, … a very expensive feat with massively long delays.</blockquote><p>Fast forward through a very long and painful journey, we eventually got the classic car restored — it looked fabulous. All those days sanding off rusted metal and repairing rust holes were worth it. Most things got assembled and tested. And guess what? It didn’t work. We had saved lots of money buying second-hand equipment that required further troubleshooting to get to work, which made the learning experience even better (you don’t learn anything when you pay for it). This required advanced knowledge of electrical and electronics (power electronics at times), as well as serial communication to troubleshoot. So be sure to understand those before attempting your conversion.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*46pKnUg7PTRadZVwJSncew.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TsFKcjgXKefJ1qzY3zYZXg.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YBu7_hRDcbh9so_1bN3fUQ.jpeg" /><figcaption>Left — New AC-55 Motor (78kW), Centre — refurbed chassis, Right — Modified driveshaft installed</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EiqLCZR3zDTXa4w3o2SftQ.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vD537I1VH8cirJnZar69wg.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qLzSsvarZFS0XuAowTfXhQ.jpeg" /><figcaption>Left — Restored cabin — so many days sanding rust, Centre — Restored grill, Right — Battery Bank, a TON of them (literally)</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ddseX7zFtWXYbjXdQkNYjA.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*r5ZvROOPz7Z-UcQI9UD9mw.jpeg" /><figcaption>Left — The motor connected to the inverter and ultimately the 336V of batteries, Right — All the electrical works</figcaption></figure><p><strong>Some business lessons</strong></p><p>We learnt a great deal while still having a lot of fun on the weekends. The one big lesson in business that this project gave me is that converting cars to electric is the worst business idea (at least for me), for the following reasons:</p><ul><li>High labor component in business idea</li></ul><p>Labor is expensive in most countries. Even in countries where labor is cheaper, you cannot easily import/export second hand cars through customs, and the transportation costs are not insignificant</p><ul><li>Hardware costs are prohibitively high at small scale (at least back then)</li><li>People are not willing to pay that price tag for a conversion, when the alternative is a new car for the same price</li><li>The margins are very low (non-existent at small scale)</li><li>A first business experience which involves low margin is a bad one (doesn’t allow for failures)</li></ul><p>So the question that I set off to answer: what is the best kind of business to grow in the current day and age? After carefully studying many startups that made it (and others that didn’t), here are some of the lessons that I learnt:</p><ul><li>Businesses with no “Why” are lost (refer you back to Simon Sinek’s book — Start with Why)</li><li>Think about the size of industry that you are targeting, consumer behaviour in that industry, and just basic economics</li><li>Businesses with a product that has a high software component have the highest ability to scale (it’s literally copy-paste and sell to the whole world)</li><li>Businesses that have a hardware product are much costlier, slower, have a high labor component, and <em>therefore need larger capital</em></li></ul><p>I’m not against hardware businesses — but to be realistic, you need what I call an <strong>“adequate</strong> <strong>funding mechanism”</strong> commensurate to idea — it’s basic physics — software is relatively just thin air, easy to scale. Hardware is physically heavy, requires work, higher labor, transport, logistics, etc.</p><p>This coincides with something I read a while back about billionaire businesses having a very low chance of failing (I can’t find the source to quote it…). Obviously, billionaires can fund business ideas adequately, often coupling it with a strong media hype to create demand (most billionaires also own/have strong influence in some form of media), therefore leading it to success!</p><p>As a final thought — note how Elon Musk only opened up his hardware businesses (Tesla, SpaceX, Neuralink) after his software business success (Paypal). Also, note that Paypal’s competitor Stripe — its founders never exited; Patrick and John Collison, now 32 and 30 years old, respectively, are two of the youngest self-made billionaires.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ed82c0fde4f8" width="1" height="1" alt=""><hr><p><a href="https://medium.com/swlh/electric-vehicle-conversion-shopping-list-and-an-expensive-lesson-in-business-ed82c0fde4f8">Electric Vehicle Conversion Shopping List and an Expensive Lesson in Business</a> was originally published in <a href="https://medium.com/swlh">The Startup</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[11 Skills For Tomorrow’s Automation Engineers]]></title>
            <link>https://medium.com/@louiskh/11-skills-for-tomorrows-automation-engineers-25ca66b92313?source=rss-7f1716cc1864------2</link>
            <guid isPermaLink="false">https://medium.com/p/25ca66b92313</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[robotics]]></category>
            <category><![CDATA[automation]]></category>
            <category><![CDATA[engineering]]></category>
            <category><![CDATA[deep-learning]]></category>
            <dc:creator><![CDATA[Louis Khalayli]]></dc:creator>
            <pubDate>Tue, 22 Sep 2020 19:46:39 GMT</pubDate>
            <atom:updated>2020-09-22T19:46:39.800Z</atom:updated>
            <content:encoded><![CDATA[<p>Automation engineers are the shakers and movers of emerging technologies. We often research, develop, and deploy technologies to achieve different goals, all under the general purpose goal of automating tasks. As a natural progression, the profession requires engineers to be cognizant of fundamentals of new often-black-box technologies and their applications, as they are introduced to industrial plants, and expectations are rising.</p><p>As production lines becoming increasingly dependent on machine learning, computer vision, and data analytics systems, here are 11 technical skills that I believe are essential for every automation engineer in tomorrow’s world:</p><p>Emerging:</p><p><strong>1.</strong> <strong>Machine Learning (and Deep Learning)</strong></p><p>This technology trend is currently spreading fast around the world. The technology has great potential to, at best disrupt, at worst marginally improve many industries; however, the limitations are usually as follows:</p><p>· On both the provider and the customer’s side as there is a massive shortage of people who have a firm understanding of the technology fundamentals and/or how it can be applied in different environments (i.e. most machine learning experts and data scientists have never worked in a production environment for long enough to understand how to apply their technology)</p><p>· The data where you want to apply machine learning is either not available, not clean or, often when available is plain wrong (company culture has low diligence on keeping data clean)</p><p>Tips for automation engineers:</p><p>· Ensure you understand the fundamentals: there are many excellent free online courses, including on coursera — it is heavily statistics-based, so skip this if you’re not a fan</p><p>· Make sure you check out free online code libraries</p><p>· If you are using ready-made solutions, the cost is usually prohibitive unless you find applications that are disruptive (i.e. bring the cost down of a certain task by &gt; 80%).</p><p><strong>2.</strong> <strong>Computer Vision</strong></p><p>Depending on the product you make, there may be great applications for computer vision. There are different types of machine or computer vision systems, and the fundamentals are as simple as those filters you see on Instagram, all the way up to Deep-Learning based systems (where every pixel is its own feature in the feature vector).</p><p>Tips:</p><p>· Read up on image processing filters</p><p>· Make sure you check out free online code libraries</p><p><strong>3.</strong> <strong>Cyber security of Industrial Control Systems (OT)</strong></p><p>This is a trend that emerged in the last 12 or so years, when hackers started targeting industrial control systems (those that control power plants, power distribution, water treatment, etc.), and started causing havoc by shutting down critical infrastructure points. See this article by the <a href="https://www.cisa.gov/news/2020/07/23/protect-operational-technologies-and-control-systems-against-cyber-attacks">Dept of CISA</a>.</p><p>The risk stems from these control systems (PLCs) being designed to be on isolated networks not connected to the internet; however, due to the requirement for remote engineering support on many of those remote plants, many of those are now connected to the internet, and thus, are -as a side effect- accessible to hackers (if you are interested, carefully check out <a href="http://www.shodan.io/">Shodan.io</a> — Caution: Do not poke around too much on Shodan — If you hack into any system, you may cause fatal damage, and may find yourself in jail. This is a website that scans IP addresses for accessible industrial control systems and IoT devices, and provides it to the public!</p><p>In the world of Operational Technology, OT, or cyber security of Industrial Control systems, you will learn lots about networks, attack vectors, vulnerabilities, threats, software and hardware, OS (Windows and Linux), antivirus, backup, patching, … It’s a world of its own.</p><p>There are still many plants around the world that follow horrific cyber security measures in their control system environments, where the risk is assumed low by management, or controlled by legal measures. Depending on where you live, there may be absolutely no awareness about OT and its importance, and hence little to no jobs in that space.</p><p>Traditional — these skills are quite well known and require no description:</p><p><strong>4.</strong> <strong>PLC and its programming</strong></p><p><strong>5.</strong> <strong>Safety control systems</strong></p><p><strong>6.</strong> <strong>Robotics — </strong>While emerging for some industries, robotics are considered traditional in the automotive industry, as the first ones were installed on production lines since the late 1960’s</p><p><strong>7.</strong> <strong>Quality systems, Management of Change, Maintenance systems, Project Management fundamentals</strong></p><p><strong>8.</strong> <strong>Mechanical dynamics fundamentals (motors, VSDs, tool control, torque control, etc..)</strong></p><p><strong>9.</strong> <strong>Electrical and Electronics fundamentals</strong></p><p><strong>10.</strong> <strong>Databases</strong></p><p><strong>11.</strong> <strong>Software fundamentals: Object oriented programming, Python, C++</strong></p><p>While one does not need to be an expert in every one of them, there can be large benefits to knowing a little about every one of them, <em>while mastering 2–3 of those </em><strong><em>across both emerging and traditional</em></strong> can truly boost your career growth.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=25ca66b92313" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>