<?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 Zohaib Saleem on Medium]]></title>
        <description><![CDATA[Stories by Zohaib Saleem on Medium]]></description>
        <link>https://medium.com/@zohaibsaleem_69651?source=rss-ddbe7d90740f------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*udnZVGx09ep0EkJ1</url>
            <title>Stories by Zohaib Saleem on Medium</title>
            <link>https://medium.com/@zohaibsaleem_69651?source=rss-ddbe7d90740f------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 23:03:07 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@zohaibsaleem_69651/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[Why My E-Commerce Venture Didn’t Work Out]]></title>
            <link>https://medium.com/@zohaibsaleem_69651/why-my-e-commerce-venture-didnt-work-out-32aea70721d0?source=rss-ddbe7d90740f------2</link>
            <guid isPermaLink="false">https://medium.com/p/32aea70721d0</guid>
            <dc:creator><![CDATA[Zohaib Saleem]]></dc:creator>
            <pubDate>Mon, 09 Dec 2024 04:31:54 GMT</pubDate>
            <atom:updated>2024-12-09T04:31:54.706Z</atom:updated>
            <content:encoded><![CDATA[<p>So, let’s talk about my e-commerce journey. You know, online stores have really taken off, right? They offer all kinds of perks that brick-and-mortar shops just can’t match. Think about it: convenience, a wider selection, and often better prices. I was convinced by all those influencer videos urging folks like me to jump into the online business pool. So, I thought, “Why not?” But here’s the kicker: I dove in without really understanding the business landscape or the strategies that make startups tick.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*X0nuKHkO9JE-bBoX0Oq6iA.jpeg" /></figure><p><strong>What Went Wrong</strong><br> <br> Now, let me tell you, running an e-commerce venture is no walk in the park. It’s basically a full-time gig, especially in those first three months. I was naive to think I could just set it up and let it run itself. Nope, not happening.</p><p><strong><em>Business Model:</em> </strong>First off, you’ve got to have a clear business model. There are different paths you can take, like drop shipping or creating your own brand. Each has its own game plan. I went for drop shipping because, well, it seemed to fit my budget and my schedule, at least, that’s what I thought. I found out later that till everything is not streamlined which takes at least three months of full time attention and spending considerable amount of money, drop shipping is also a full time job. <br> <br><em> </em><strong><em>Marketing:</em></strong> I didn’t define my marketing strategies. In a place like Pakistan, understanding your target audience is crucial. It’s not just about throwing up a website and hoping people show up. In Pakistan, a combo of digital tools like ads on Meta or TikTok and good old-fashioned methods, like teaming up with influencers or hosting local events, often works wonders. Without a clear strategy, my marketing ended up being all over the place and it just didn’t hit the mark.<br> You’ve got to know how to reach them effectively.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aU0HPCKZFUHazbgeEBO87Q.jpeg" /></figure><p><strong><em>Middleman woes in drop shipping: </em></strong>At first, drop-shipping seemed like a low-risk choice, but depending on third-party suppliers? That’s a whole other story. It can lead to mixed product quality, shipping delays, and squeezed profit margins. If you’re looking to import products, you need some serious capital. Cutting out the middleman sounds great, but it means buying in bulk, which is a huge upfront cost. On the flip side, local retailers will sell you products one at a time, but that really cuts into your profit margins. Key is to keep searching for the winning product which qualifies basic attributes such as price, uniqueness, quality and numbers available with vendor. Generally drop shipping winning products have very short life span and in this short span all the processes of acquiring, marketing and delivery should be streamlined to earn profits.<br> <br> <strong><em>Quantity versus quality:</em></strong> Finding that sweet spot between product quality and quantity is a huge aspect of e-commerce success. It’s tempting to sell cheap, low-quality products in bulk because they move quickly. But here’s the catch, customers might buy once, but they won’t come back. On the other hand, if you sell high-quality items at a premium price, you’ll find that only a few people are willing to buy. In Pakistan, offering products that are of decent quality at fair prices might just work. But sourcing those can be a headache, high-quality usually means higher prices, and cheap stuff might not cut it. It’s a tricky balance.<br> <br><em> </em><strong><em>Pricing strategies:</em></strong> This can make or break your business too. The e-commerce landscape is super competitive, and that often leads to price wars. I got caught up in trying to undercut my rivals, which ended up hurting my profit margins without really enhancing the customer experience. In the drop shipping game, everyone’s sourcing from the same local wholesalers, which leads to price wars. You’ve got big retailers undercutting prices, making it nearly impossible for smaller sellers to compete. That leaves you with little room to maneuver when it comes to marketing and selling your products. If I’d thought about dynamic pricing; adjusting prices based on demand, competition, and seasonal trends, I could have stayed competitive while still making a profit. But in such a cutthroat market, managing that dynamic pricing thing would have needed constant tweaking to balance decent profits with prices that attract customers.<br> <br> <strong><em>Delivery services:</em></strong> They’ve improved, sure, but getting stuff to rural areas? That’s a whole different ball game. Many delivery companies won’t go to places without clear addresses or postal codes. In rural areas most addresses are just landmarks, which makes things complicated. Imagine trying to coordinate 20 deliveries a day where everyone has to meet at a landmark. Stressful, right?<br> <br><em> </em><strong><em>Tracking Orders:</em></strong> Order tracking and managing returns is a nightmare, especially in Pakistan where cash-on-delivery is the norm. It leads to a lot of hiccups — think abrupt cancellations or returns that are a headache for both retailers and delivery services. Plus, for every failed delivery, you get hit with a fee.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gIPekEx0k7DwIoOZfJLxXQ.jpeg" /></figure><p><strong>Conclusion</strong><br> So, in the end, e-commerce isn’t just about having a website; just like any other business, it’s about planning and strategy. It’s time-consuming and really demands your full attention, especially at the start. With so many new players and big corporations in the mix, small retailers face an uphill battle. But, hey, it’s not impossible! If you nail your marketing and sourcing strategies, you can make it work. Just be sure to spend a good three months learning about startup strategies, the current business climate, and figuring out what went wrong in my case.</p><p><strong>Learning</strong> from my past challenges and driven by a desire to improve, I immersed myself in extensive research and analysis of successful startup models. This journey of growth enabled me to provide strategic support to multiple e-commerce ventures, helping them navigate their formative stages with confidence and clarity</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=32aea70721d0" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Relational Databases as I/O Tags in Distributed Control Systems]]></title>
            <link>https://medium.com/@zohaibsaleem_69651/relational-databases-as-i-o-tags-in-distributed-control-systems-d4b56007f7ce?source=rss-ddbe7d90740f------2</link>
            <guid isPermaLink="false">https://medium.com/p/d4b56007f7ce</guid>
            <dc:creator><![CDATA[Zohaib Saleem]]></dc:creator>
            <pubDate>Tue, 05 Nov 2024 06:20:32 GMT</pubDate>
            <atom:updated>2024-11-05T06:20:32.635Z</atom:updated>
            <content:encoded><![CDATA[<p>Distributed control systems (DCS) are the backbone of modern industrial automation, managing complex processes across a wide range of industries.</p><p>DCS systems can be considered as central nervous system of modern industries. Real time data from the sensors, analyzers, actuator feedbacks etc. installed in the field is received and based on this data, actions defined in logic are implemented, displayed on HMI, user input are registered and subsequent actions are taken through actuators, to control processes.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*X3Hp94u5Hyg6XsTv4hjWuQ.jpeg" /><figcaption>Typical DCS architecture</figcaption></figure><p>Central to these systems is the concept of Input/Output tags or accounts, which represent digital or analog values connected to physical devices such as sensors and actuators. These tags are user or system designer generated and classified into soft and hardware (I/O) tags. Soft tags are normally used as connections or registers between different logic blocks, whereas I/O tags represent actual sensor or actuator attached in the field. All of these tags are treated as tables with multiple records, some of which are:</p><p>· Classification of tag</p><p>· Tag identification</p><p>· Point where this tag is connected</p><p>· Tag description</p><p>· Type of hardware connected</p><p>· Any mathematical conversions required</p><p>· Scaling of tag</p><p>· Where this Tag will be used in logic</p><p>· Engineering units of this tag</p><p>· Scan time of the tag or frequency of tag</p><p>· Details about trend of tag</p><p>· Alarm limits of tag</p><p>and many other attributes. Typically, these tags were managed within the DCS system. However, with the complexity of industrial operations and data-driven decision making, there is a growing trend to use relational databases to store and manage I/O tags. Using data normalization process, these records are split into dimensional tables based on relevancy.</p><p><strong>Why do I/O tags involve database?</strong></p><p>A typical DCS system have more than 5000 soft or hardware tags with all the information listed above. To manage queries and logic sequencing from these tags in real time, databases offer streamlined and efficient ways of handling these queries. Other advantages are:</p><p>· Central Data Storage: Relational data storage provides central data storage for all I/O tags, enabling efficient data management and retrieval.</p><p>· Data Integrity and Consistency: Data security and normality constraints ensure data stability and prevent inconsistencies and errors.</p><p>· Scalability: Relational databases can easily handle large enterprise workloads with thousands of I/O tags.</p><p>· Query capabilities: SQL can perform complex queries based on historical data and time.</p><p>· Security and access control: Relational data centers provide strong security measures to protect sensitive corporate data.</p><p>· Integration with other systems: Databases can be integrated with other systems, such as reporting systems, analytical platforms, and business intelligence tools.</p><p>Let’s dwell in how relational databases are implented in DCS:</p><p><strong>1. <em>Account details and configuration:</em></strong></p><p>· Create database tables to store tag information: tag name, description, data type (digital or analog), engineering units, alarm limits, etc.</p><p>· Define relationships between tags (e.g., parent-child relationships for classification).</p><p><strong>2.<em> Real-time Database:</em></strong></p><p>· Establish a continuous flow of data from the DCS to the database through intervention or centralization.</p><p>· Use database triggers or stored procedures to dynamically update code values.<br><strong> 3.<em> Historical Data Storage:</em></strong></p><p>· Configure the database to save historical data on a regular basis (e.g., hourly, daily, or weekly).</p><p>· Implement data retention policies to better manage storage space.</p><p><strong>4. <em>Data Analysis and Reporting:</em></strong></p><p>· Analyze historical data using SQL queries and generate key performance indicator (KPI) reports.</p><p>· Connect business intelligence tools for advanced data visualization and reporting.</p><p><strong>Benefits of Using a Relational Database</strong></p><p>The improvement of data quality in DCS is achieved by using a database with strict validation rules, which guarantees the accuracy and consistency of all data. This reliable database improves decision-making because it provides easy access to historical and live information, enabling informed and strategic choices based on data.</p><p>Relational database simplify tasks through automated data collection and analysis, significantly reducing manual tasks and increasing operational efficiency, allowing resources to be allocated to more important tasks.<br> <br> Security is a critical function that requires the ability to protect the company’s sensitive information and protect the integrity and confidentiality of data. Relational databases seamlessly integrates with other systems to provide a comprehensive and integrated picture of organizational operations. This integrated approach promotes a smoother workflow, makes it easier to monitor and improve all aspects of your work.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/300/1*5Sveq4T5v-tyl0DrKpGAZg.png" /><figcaption>DCS controller with I/O cards (Courtesy Emerson Delta V)</figcaption></figure><p>By leveraging relational databases for I/O tag management, industrial organizations can unlock the full potential of their DCS systems. The benefits of improved data quality, scalability, and integration capabilities make this approach a compelling choice for modern industrial automation.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d4b56007f7ce" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Harnessing AI & ML for Distributed Control Systems (DCS) in Industry]]></title>
            <link>https://medium.com/@zohaibsaleem_69651/harnessing-ai-ml-for-distributed-control-systems-dcs-in-industry-e85ae44ee2e3?source=rss-ddbe7d90740f------2</link>
            <guid isPermaLink="false">https://medium.com/p/e85ae44ee2e3</guid>
            <dc:creator><![CDATA[Zohaib Saleem]]></dc:creator>
            <pubDate>Mon, 28 Oct 2024 12:37:19 GMT</pubDate>
            <atom:updated>2024-10-28T12:37:19.762Z</atom:updated>
            <content:encoded><![CDATA[<p>In recent years, Artificial Intelligence (AI) and Machine Learning (ML) have dramatically transformed the landscape of Distributed Control Systems (DCS) in industrial settings. Traditionally, DCS systems were designed to monitor and control complex processes, but as industries grow more data-intensive, AI and ML provide a powerful upgrade to these systems. However, there are few factors which need to be carefully deliberated before AI &amp; ML implementation, such as processing load on DCS controllers, cyber security additional costs and requirement of skilled workforce. But before dwelling in the challenges let’s explore how these technologies enhance DCS by improving efficiency, safety, and operational insight.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kTN1qy0KsY-puABnY3WVyA.jpeg" /></figure><p>1. Predictive Maintenance and Reduced Downtime</p><p>One of the greatest benefits AI and ML bring to DCS is predictive maintenance. By analyzing historical and real-time data from machinery and sensors, machine learning algorithms can detect patterns that precede equipment failure. This enables the DCS to notify operators before a malfunction occurs, allowing for planned maintenance rather than reactive repairs. Not only does this minimize unplanned downtime, but it also prolongs equipment life and reduces maintenance costs.</p><p>For example, in an oil refinery, predictive maintenance on pumps and compressors can prevent costly production halts by identifying issues like vibration anomalies or temperature increases long before they escalate into a failure.</p><p>2. Enhanced Process Optimization</p><p>AI and ML can autonomously analyze and optimize complex processes that were previously managed through manual adjustments. Using data collected from sensors, the DCS can apply machine learning models to predict optimal operating conditions. By adjusting parameters like flow rates, temperature, and pressure, the system ensures peak performance, improving yield and reducing energy consumption.</p><p>This optimization is helpful in chemical processing, where AI-driven optimization can precisely control reaction temperatures and flow rates, enhancing production efficiency and reducing waste.</p><p>3. Improved Anomaly Detection</p><p>Anomalies in industrial processes often signal critical issues, but manual identification is challenging due to the volume and complexity of data. Machine learning algorithms trained on historical process data can identify even subtle anomalies that may escape human detection. This empowers the DCS to issue early alerts and prompt corrective actions.</p><p>For example in power generation industries, an ML algorithm can detect deviations in turbine speed or gas composition, alerting operators to potential safety hazards before they impact operations.</p><p>4. Advanced Control Strategies with Reinforcement Learning</p><p>Reinforcement learning, a subset of machine learning, allows AI systems to learn optimal control strategies through trial and error within simulated environments. In a DCS, reinforcement learning can optimize the control strategy by continuously adjusting parameters to achieve specific objectives, like maximizing output or minimizing fuel consumption, all while maintaining safety limits.</p><p>This is particularly helpful in water treatment plants, where reinforcement learning can dynamically adjust chemical dosing and pump speeds to maintain water quality while reducing energy usage.</p><p>5. Real-Time Data Insights and Decision Support</p><p>AI and ML provide real-time insights by continuously analyzing sensor data across a DCS network. These insights are often visualized in dashboards, helping operators make informed decisions swiftly. Decision-support systems powered by AI can also recommend specific actions, helping operators navigate complex situations more effectively and with greater confidence.</p><p>In manufacturing, AI-powered dashboards that aggregate sensor data from the entire production line enable operators to track performance in real-time, quickly respond to anomalies, and maintain efficient production flows.</p><p>6. Increased Safety and Compliance</p><p>In industries like oil and gas or chemicals, safety is paramount. AI-powered DCS can monitor safety-critical equipment, detect unsafe operating conditions, and even initiate automated shutdowns when necessary. By maintaining detailed records of operations and identifying trends, AI also supports compliance with industry regulations, which is crucial in avoiding costly penalties and ensuring safe working conditions.</p><p>In oil &amp; gas processing units, AI models monitoring hazardous material tanks can detect leaks early by analyzing pressure and gas levels, automatically triggering containment protocols to protect personnel and the environment.</p><p>7. Energy Efficiency and Sustainability</p><p>AI-driven DCS helps industries achieve sustainability goals by optimizing energy consumption. Machine learning models can pinpoint areas of inefficiency, such as excessive power use in certain equipment, and recommend adjustments that reduce energy waste. This is particularly valuable in energy-intensive industries like manufacturing and mining.</p><p>AI can optimize boiler operations in a power plant by adjusting fuel input, airflow, and temperature to ensure complete combustion, thereby reducing emissions and improving fuel efficiency.</p><p>Challenges and Considerations</p><p>While the benefits of integrating AI and ML with DCS are clear, successful implementation requires attention to several factors. These factors are to be carefully studied before moving forward for implementation of ML&amp;AI:</p><ul><li>Controller Processing Optimization: Normally a DCS controller is sequentially doing three or four major tasks depending upon the DCS manufacturer. These tasks are logic execution, execution of watch dog timer and controller uplink/downlink communications. Real time data extraction for AI &amp; ML puts extra burden of communications on controllers. Generally controller processing load should not exceed 50% of its total resource time. There are ways to avoid this problem, but these entail extra costs.</li><li>Data Quality: High-quality data is essential for accurate model training. Inconsistent or noisy data can lead to unreliable predictions.</li><li>Cybersecurity: As DCS becomes more interconnected with AI systems, robust cybersecurity measures are crucial to prevent unauthorized access and protect sensitive information.</li><li>Skilled Workforce: Integrating AI into industrial systems demands expertise in both industrial processes and data science. Training operators and engineers is key to leveraging AI effectively.</li></ul><p>The integration of AI and ML into Distributed Control Systems marks a new era of smart, data-driven industrial control. By enabling predictive maintenance, enhancing process optimization, detecting anomalies, and improving safety, AI-powered DCS systems not only increase operational efficiency but also support sustainability and safety. As industries continue to embrace these advancements, AI-driven DCS will be fundamental to the future of industrial automation, paving the way for smarter, safer, and more sustainable operations.</p><p>Industries looking to remain competitive and resilient should consider adopting AI and ML solutions for their DCS, not only to optimize current processes but also to future-proof their operations in an increasingly data-centric world.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e85ae44ee2e3" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>