[{"content":"🔗 Paper on which the article is based written by Mohammed Bilal Ahmed and Nicolas Perez\n✏️ Introduction Mapping algorithms are foundational to geospatial applications, enabling efficient navigation, urban planning, and real-time routing. This peer-reviewed paper explores graph-based algorithms, such as Dijkstra’s and A* (A-star), for processing spatial data in mapping tasks. The authors aim to evaluate these algorithms’ efficiency in handling large-scale geospatial datasets, focusing on computational complexity, memory usage, and adaptability to dynamic environments like real-time traffic updates. The study uses graph theory, where locations are represented as nodes and paths as edges, to address challenges in creating accurate and scalable maps for applications like air-traffic control and autonomous vehicle routing.\nWhat did the authors assess and find? The authors tested Dijkstra’s algorithm, A*, and a modified Bellman-Ford algorithm on urban road network datasets ranging from 1,000 to 1,000,000 nodes, sourced from OpenStreetMap. They simulated scenarios like shortest-path routing and dynamic obstacle avoidance, with 20 GIS professionals providing feedback on algorithm outputs in practical mapping tools.\nThe results showed that A* outperformed Dijkstra’s, reducing computation time by up to 40% in dense urban networks due to its heuristic approach. Dijkstra’s algorithm was more reliable for optimal paths in sparse networks but struggled with memory demands for larger datasets. The modified Bellman-Ford algorithm excelled in dynamic rerouting but was 25% slower than A* on average. Interestingly, A*’s performance dipped when heuristics were not tailored to specific map types (e.g., indoor vs. outdoor). Preprocessing geospatial data, such as simplifying road networks, significantly boosted all algorithms’ efficiency.\nLimitations of the paper The paper highlights several limitations. The use of synthetic traffic scenarios may not fully reflect real-world complexities, such as unpredictable road closures. The sample size of 20 GIS professionals is relatively small, potentially limiting the generalizability of user insights. The study focused exclusively on graph-based algorithms, omitting emerging machine learning-based approaches that could enhance mapping tasks. Additionally, the datasets were primarily urban, so performance in rural or mixed terrains remains untested. These constraints suggest caution when applying the findings to diverse geospatial contexts.\nThe contributions of the paper to the scientific community This paper offers a robust comparison of graph-based algorithms, providing practical guidance for developers building geospatial systems like navigation tools or air-traffic control software. By benchmarking performance on real-world datasets, it helps engineers select algorithms suited to specific mapping tasks, such as path-planning for autonomous vehicles. The emphasis on tailoring A* heuristics underscores a key consideration for optimizing routing efficiency. As geospatial technologies advance—particularly with 3D mapping and IoT integration—this work serves as a foundation for future research into hybrid algorithms combining graph theory with artificial intelligence, ensuring mapping systems remain scalable and responsive.\n","permalink":"https://mdbilalahmed.netlify.app/blog/paper-on-mapping/","summary":"🔗 Paper on which the article is based written by Mohammed Bilal Ahmed and Nicolas Perez\n✏️ Introduction Mapping algorithms are foundational to geospatial applications, enabling efficient navigation, urban planning, and real-time routing. This peer-reviewed paper explores graph-based algorithms, such as Dijkstra’s and A* (A-star), for processing spatial data in mapping tasks. The authors aim to evaluate these algorithms’ efficiency in handling large-scale geospatial datasets, focusing on computational complexity, memory usage, and adaptability to dynamic environments like real-time traffic updates.","title":"Summary: Graph-based Spatial Data Processing for Mapping Applications"},{"content":"🔗 Link to Paper on which the article is based written by Mohammed Bilal Ahmed\nIntroduction Throughout the last few semesters, I\u0026rsquo;ve noticed how students have a hard time understanding complex projects and the expected deliverables associated with them. Creating large projects from scratch is daunting to a beginning programmer. So, in the spring of 2023 I along with Alex Ng and Seth Rogers worked on a way to help students break down complex projects and teach them how to code using different planning methods. Pseudo-code provides a way to create and understand code and allows students to break down these larger problems into more manageable sections. The goal was that students who took this workshop would have a more structured approach to their projects and would allow students to break down these larger problems into more manageable sections.\nMethod and Implementation We choose a workshop format to teach students how to use pseudo-code. The workshop was broken down into 3 parts. The first part was a lecture on pseudo-code and how to use it. The second part was a group activity where students would work together to create a pseudo-code for a given problem. The third part was a lab where students would work on their own to create pseudo-code for a given problem. We choose a workshop format because it was the most familiar to the students. The students were used to attending workshops and labs for their classes. We also choose a workshop format because it allowed us to teach students how to use pseudo-code in a more hands-on way. The students were able to practice using pseudo-code in a group setting and on their own.\nLimitations of the paper The workshop was only tested on a small group of students. We only had the oppurtunirty to test the workshop on 10 students. A larger sample size would have been beneficial. The workshop was also only tested on students who had some programming experience. We did not test the workshop on students who had no programming experience. This would have been beneficial because it would have allowed us to see if the workshop was effective for students who had no programming experience.\nNext Steps We hope to implement this lesson on pseudo-code inside of the curriculum at universities since it is so important to the understanding of programming. The next step would be to introduce this workshop into the curriculum through the professors and eventually teach TAs how to teach this information themselves. The link to the paper can be found at the start of the page.\n","permalink":"https://mdbilalahmed.netlify.app/blog/coding-workshop/","summary":"🔗 Link to Paper on which the article is based written by Mohammed Bilal Ahmed\nIntroduction Throughout the last few semesters, I\u0026rsquo;ve noticed how students have a hard time understanding complex projects and the expected deliverables associated with them. Creating large projects from scratch is daunting to a beginning programmer. So, in the spring of 2023 I along with Alex Ng and Seth Rogers worked on a way to help students break down complex projects and teach them how to code using different planning methods.","title":"A new look into CS education Pt 1."},{"content":"🔗 Paper on which the article is based written by Mohammed Bilal Ahmed and Nicolas Perez\nIntroduction Fitts law: A speed and accuracy model of human movement developed from research in a machine system for air traffic. The goal of this paper was to use this very law and try to fit it into 2-dimensional tasks. When the authors had examine the law from this perspective. They had to consider many different factors such as the angle at which the target might be and so on. There were different formulas developed beforehand but the Shannan formula was preferred for the basis of this experiment. Different factors such as the size and shape of the model were taken into consideration when calculating the movement time(i.e., time required to touch the target)\nWhat did the authors assess and find? They assessed the time taken by the users to select a target at different angles. Each angle and general distance were categorized into different groups. There were twelve computer literature subjects that served as volunteers. There were 9 male and 3 female volunteers. The results from these tests were then compared. The SMALLER-OF and W’ models were very close in their movement time. While on the other hand the SMALLER-OF vs STATUS QUO and W’ vs STATUS QUO had a huge gap in their movement time. The SMALLER-OF model highlighted that Movement time depends upon pointing distance more as the target gets larger. An interesting observation under the STATUS quo model was that generally, one would expect large size letters and alphabets should have a significantly shorter time to select but then it turns out that the results or prediction did not match this. A major observation that was found was that it was difficult to compare these different models as in some such as W’ it was assumed that the subject moved towards the center of the target. Perhaps the biggest factor was comparing these different models when the parameters themselves were not the same. So the comparison did not make sense other than the final movement time.\nLimitations of the paper The paper itself talks about some of the limitations of the proposed models. While some models are easier to use, such as the Status Quo model, they can lead to inaccurate data. The other models, which account more for height and angle of approach, can be harder to apply in a more general sense, such as the smaller-of model being limited to rectangular objects. Notably what isn’t discussed as a limitation with regards to their experiment, is that the sample size for the experiment is fairly small: 12. In addition, the subjects of the study were computer-literate volunteers. However, the volunteers' backgrounds are less of a problem since the focus of the study was to fix Fitt’s law in two dimensions, since it could with certain configurations produce a “negative” difficulty value.\nThe contributions of the paper to the scientific community This paper has a clear scientific contribution through its study of ways of modifying Fitt’s law into a more useful form for two-dimensional environments. Since most of our User Interfaces are two-dimensional, this is important information to know. In addition, instead of just one model, it provides and shows the result of several models and how accurate they are. It’s important that the fundamentals we rely on stay up to date as technology changes, and this paper does that for Fitt’s law to bring it to two dimensions. Work in this area should continue as User Interfaces change and grow, such as in three-dimensional environments as Virtual and Augmented Reality devices grow in popularity.\n","permalink":"https://mdbilalahmed.netlify.app/blog/aws-saa-certification/","summary":"🔗 Paper on which the article is based written by Mohammed Bilal Ahmed and Nicolas Perez\nIntroduction Fitts law: A speed and accuracy model of human movement developed from research in a machine system for air traffic. The goal of this paper was to use this very law and try to fit it into 2-dimensional tasks. When the authors had examine the law from this perspective. They had to consider many different factors such as the angle at which the target might be and so on.","title":"Summary: Extending Fitts' law to two-dimensional tasks"},{"content":"🔗 Paper on which the article is based written by Mohammed Bilal Ahmed and Nicolas Perez\n✏️ Introduction The authors of the paper focused on the heuristic analysis of user interfaces. They wanted to test and see what could be the best way to identify problems in a user interface. Although heuristic analysis is not the only way. It is by far the most practical when compared to other options.\nWhat did the authors assess and find? To test its effectiveness the authors of the paper conducted 4 experiments with different parameters. Each experiment had either a different set of evaluators or a different set of software that the team of evaluators had to assess. The first and the second experiment did not have live working software to test. Furthermore, these two experiments had different sets of evaluators. The first experiment had a team of university students with some educational background in heuristics and the second experiment was done by working industry professionals. However, the last two experiments had the same sets of evaluators but different live working software. Through these experiments, the author was able to study the effectiveness of a heuristic analysis of a user interface.\nLimitations of the paper The authors found that after a certain point. Adding more evaluators did not have a significant impact. Although this is dependent on the complexity of the software as well. In any case, none of the evaluators were able to find 100% of the problem. In Fact, most problems found by the team of evaluators were in experiment one and that too was only 51%. Another interesting point that was highlighted in the experiment was the fact that heuristic analysis works best when one aggregates a team\u0026rsquo;s results. This also ties in with the idea of what one defines as a “usability” problem to begin with. Different evaluators may have different ideas about what constitutes a usability problem. Aggregating results in such cases helps one see issues in a more holistic light. This then leads to the limitations of this paper. The paper does have some limitations in the demographics that it reaches. The two demographics that are covered are college students who recently had a lecture on evaluation heuristics and industry professionals. The paper itself points to this limitation by making a note that usability experts would be better at identifying problems through heuristic analysis. In addition, three of the four tests in the paper were done on only one of the covered demographics: college students. However, the data from the industry professionals fits the same trends as the data from students, which points to there being little correlation between the demographics of the evaluators and the conclusions on aggregate analysis. More data should be collected to be able to state that for sure.\nThe contributions of the paper to the scientific community While the ideas the paper covers may seem rather simple, the paper presents a large positive contribution to both the academic community and the industry as a whole. Everything in life is built off other things, and this paper functions as a foundation to work off of. It uses experimental data to show just how many resources should be allocated to test the functionality of user interfaces without wasting both time and money. It also provides recommendations on how best to analyze user interfaces and find problems.\n","permalink":"https://mdbilalahmed.netlify.app/blog/kindle-to-notion/","summary":"🔗 Paper on which the article is based written by Mohammed Bilal Ahmed and Nicolas Perez\n✏️ Introduction The authors of the paper focused on the heuristic analysis of user interfaces. They wanted to test and see what could be the best way to identify problems in a user interface. Although heuristic analysis is not the only way. It is by far the most practical when compared to other options.","title":"Summary: Heuristic evaluation of user interfaces."},{"content":"🔗 Paper on which the article is based written by Mohammed Bilal Ahmed and Nicolas Perez\n✏️ Introduction Human Computer interaction can be a lot more than just making convenient systems that make tasks and work easier. This is shown by the paper we’ll be analyzing: A Brain-Computer Interface and Augmented Reality Neurofeedback to Treat ADHD: A Virtual Telekinesis Approach. The authors of the paper use previous research to describe and justify a new form of Neurofeedback Therapy for kids with ADHD. This new form of NFT uses AR technology as opposed to two dimensional interfaces from the past.\nWhat did the authors assess and find? The new gamified therapy the authors present functions by using a device to measure the type of brainwave the user\u0026rsquo;s brain produces. By prior research on the authors’ part, they note that people with ADHD produce more “Theta” brain waves and less “Beta” brain waves than neurotypical people. The aim of this new interface, as with previous methods of Neurofeedback Therapy, is to increase the number of “Beta” brainwaves the user produces. The authors note that in the past this form of therapy has issues with users committing to it due to the number of sessions it requires, and the lack of engaging feedback. This new interface was created to solve that issue with its AR interface. In particular, this interface uses a balloon inflating and a spoon bending in correlation to the user’s focus.\nLimitations of the paper As this is an “idea” paper, there isn’t much data the authors collected themselves. The data they collected was used in order to help them correctly use the data collected by the headset the users have to wear. Authors accounted for differences between individuals by using an algorithm that adjusts to the individual, this does require a baseline examination however. That baseline examination also accounts for differences created by the user moving their head or blinking. From feedback, they assessed that the more realistic the models and animations they had the better the interface was. The paper is not an experimental paper rather it is more of a white paper. It does an excellent job at introducing AR and VR in relation to neurofeedback. However one of the biggest drawbacks of the paper was the fact that it did not follow through with field experiments for many of the ideas that were presented. Thus, one could not verify the effectiveness of the ideas that were presented.\nThe contributions of the paper to the scientific community The paper not only explored a new field of research but also provided enough evidence for other researchers to potentially further the study. The paper’s biggest contribution was the fact that it explored most if not all the possibilities for using AR and VR for Neurofeedback. The analysis of the feasibility and success rate of each AR and VR method mentioned are a subject of a study in itself. That is what the paper aimed at presenting i.e. it did not aim at evaluating each method\u0026rsquo;s practical success rate rather it aimed at presenting all the available options. All in all this paper is only a first step in the field of Neurofeedback and AR/VR technology. There are many more steps needed before one can implement this in a real clinical case.\n","permalink":"https://mdbilalahmed.netlify.app/blog/face-landmarks-detection/","summary":"🔗 Paper on which the article is based written by Mohammed Bilal Ahmed and Nicolas Perez\n✏️ Introduction Human Computer interaction can be a lot more than just making convenient systems that make tasks and work easier. This is shown by the paper we’ll be analyzing: A Brain-Computer Interface and Augmented Reality Neurofeedback to Treat ADHD: A Virtual Telekinesis Approach. The authors of the paper use previous research to describe and justify a new form of Neurofeedback Therapy for kids with ADHD.","title":"Summary: Brain-Computer Interface and Augmented Reality Neurofeedback to Treat ADHD: A Virtual Telekinesis Approach"},{"content":"🔗 Link to the Project This is an interactive pathfinding visualizer built with React and JavaScript, showcasing classic algorithms like Dijkstra’s in a grid-based environment. Users can set start/end points, add walls, and watch the algorithm find the shortest path in real time.\nThis project reflects my interest in spatial computing and routing logic — areas I’m excited to explore professionally.\n🔗 Live Demo View the project live here: https://path-mapper.netlify.app/\n🧑‍💻 Run Locally git clone https://github.com/m-bilal-ahmed/Visualizing-Paths.git cd pathfinding-visualizer npm install npm start ","permalink":"https://mdbilalahmed.netlify.app/projects/visualizing-paths/","summary":"🔗 Link to the Project This is an interactive pathfinding visualizer built with React and JavaScript, showcasing classic algorithms like Dijkstra’s in a grid-based environment. Users can set start/end points, add walls, and watch the algorithm find the shortest path in real time.\nThis project reflects my interest in spatial computing and routing logic — areas I’m excited to explore professionally.\n🔗 Live Demo View the project live here: https://path-mapper.netlify.app/","title":"Visualizing Paths: interactive pathfinding visualizer"},{"content":"🔗 Link to the Project We designed a VR application keeping user design in mind. We focused on Human Computer interaction to maximize user experience. We created multiple personas to help us better understand user needs. A more detailed discription can be found on the link above.\n","permalink":"https://mdbilalahmed.netlify.app/projects/obsidian-publish-github-action/","summary":"🔗 Link to the Project We designed a VR application keeping user design in mind. We focused on Human Computer interaction to maximize user experience. We created multiple personas to help us better understand user needs. A more detailed discription can be found on the link above.","title":"ACM UF's Virtual Envoirment"},{"content":"🔗 Link to the Project Description We designed an application keeping user design in mind. We focused on Human Computer interaction to maximize user experience. We created multiple personas to help us better understand user needs. A more detailed discription can be found on the link above.\n","permalink":"https://mdbilalahmed.netlify.app/projects/kindle-to-notion/","summary":"🔗 Link to the Project Description We designed an application keeping user design in mind. We focused on Human Computer interaction to maximize user experience. We created multiple personas to help us better understand user needs. A more detailed discription can be found on the link above.","title":"Gator Carpool"},{"content":"Description [-] Built a Slack chatbot from scratch to assist a startup’s Slack channel of 300+ users. Used Python to code the bot using Slack’s Web API. [-]\tBot can also handle events allowing users to subscribe to events. Used Flask to retrieve and handle these events.\n","permalink":"https://mdbilalahmed.netlify.app/projects/automated-image-captioning/","summary":"Description [-] Built a Slack chatbot from scratch to assist a startup’s Slack channel of 300+ users. Used Python to code the bot using Slack’s Web API. [-]\tBot can also handle events allowing users to subscribe to events. Used Flask to retrieve and handle these events.","title":"Slack Assistant Bot"},{"content":"Discription  Utilized C# and .NET MVC architecture with Razor syntax to develop a responsive inventory application. Enhanced data tables with React and jQuery for real-time updates, reducing data retrieval time by 30% and improving page load speed by 25% through optimized Ajax requests. Optimized PL/SQL code for Oracle SQL, integrating stored procedures with .NET using Dapper, reducing query execution time by 20%. Developed and integrated RESTful API within a microservices architecture, improving data handling efficiency and application scalability. Automated HTML to PDF conversion using the HiQPDF .NET Library, reducing processing time by 70%, and applied Bootstrap for responsive front-end design.  ","permalink":"https://mdbilalahmed.netlify.app/experience/r+l/","summary":"Discription  Utilized C# and .NET MVC architecture with Razor syntax to develop a responsive inventory application. Enhanced data tables with React and jQuery for real-time updates, reducing data retrieval time by 30% and improving page load speed by 25% through optimized Ajax requests. Optimized PL/SQL code for Oracle SQL, integrating stored procedures with .NET using Dapper, reducing query execution time by 20%. Developed and integrated RESTful API within a microservices architecture, improving data handling efficiency and application scalability.","title":"Software Engineer / .NET Developer"},{"content":"Discription  Built a feature to allow users to update the Unix shell used on their local computer. Developed the API call to the application, and stored the data efficiently in Postgres, SQL, and MySQL databases. Developed an automated email response system using Django to enhance the account creation process by notifying users of their account status. Resulted in a 60% reduction in traffic. Utilized Agile principles to redesign and optimize an application\u0026rsquo;s front end to enhance user experience, reduce clutter, and increase account registrations by 50%. Automated the build, test, and deployment of containerized applications using Docker, resulting in a 30% reduction in deployment time.  ","permalink":"https://mdbilalahmed.netlify.app/experience/cise-it/","summary":"Discription  Built a feature to allow users to update the Unix shell used on their local computer. Developed the API call to the application, and stored the data efficiently in Postgres, SQL, and MySQL databases. Developed an automated email response system using Django to enhance the account creation process by notifying users of their account status. Resulted in a 60% reduction in traffic. Utilized Agile principles to redesign and optimize an application\u0026rsquo;s front end to enhance user experience, reduce clutter, and increase account registrations by 50%.","title":"Software Engineer"},{"content":"Discription  Led a development team using Python to build a machine learning model using Mask - RCNN classification that would recognize and then classify damaged and undamaged cars based on user-provided images. Partnered with a developer to implement a RESTful API in Flask to enable users to upload images. Used Scikit-learn and Scikit-image, for machine learning, and image processing respectively. Drove a unit test effort to improve the accuracy of the model from 56% to 90%. Led process to deploy the application onto Azure. Used Azure Cognitive Services for the ML models and Azure Custom Vision for improving the accuracy of processing images by up to 99.9%.  ","permalink":"https://mdbilalahmed.netlify.app/experience/buyerassist/","summary":"Discription  Led a development team using Python to build a machine learning model using Mask - RCNN classification that would recognize and then classify damaged and undamaged cars based on user-provided images. Partnered with a developer to implement a RESTful API in Flask to enable users to upload images. Used Scikit-learn and Scikit-image, for machine learning, and image processing respectively. Drove a unit test effort to improve the accuracy of the model from 56% to 90%.","title":"Software Engineering Intern"},{"content":"Discription  Developed sound mitigation tests in Java to limit C-130 aircraft engine sound waves. Mitigation model improved the maintenance testing time of engines by up to 65% Interviewed over 60+ military personnel and analyzed the final prototype’s cost.  ","permalink":"https://mdbilalahmed.netlify.app/experience/usaf/","summary":"Discription  Developed sound mitigation tests in Java to limit C-130 aircraft engine sound waves. Mitigation model improved the maintenance testing time of engines by up to 65% Interviewed over 60+ military personnel and analyzed the final prototype’s cost.  ","title":"Software Engineering Intern"},{"content":"Description  Taught 150+ students enrolled in Calculus-I course and organized weekly student-led sessions. Held weekly office hours and explained analytical concepts to students  ","permalink":"https://mdbilalahmed.netlify.app/experience/tumunich/","summary":"Description  Taught 150+ students enrolled in Calculus-I course and organized weekly student-led sessions. Held weekly office hours and explained analytical concepts to students  ","title":"Teaching Assistant, Calculus-1"},{"content":"Preview  Description  Co-founded a startup that helped rescue operators locate people trapped under rubble after a natural disaster occurs. Built a device \u0026amp; coded it in C++ to Increase the efficiency to locate people by 85%. Device used a motion, heat, and other sensors for an optimal location. Initiated a unit test effort to improve accuracy of location by 10% Used Python (folium \u0026amp; Pandas) to build an interactive UI and map for rescue operators to track people in real time.  ","permalink":"https://mdbilalahmed.netlify.app/experience/origin-health/","summary":"Preview  Description  Co-founded a startup that helped rescue operators locate people trapped under rubble after a natural disaster occurs. Built a device \u0026amp; coded it in C++ to Increase the efficiency to locate people by 85%. Device used a motion, heat, and other sensors for an optimal location. Initiated a unit test effort to improve accuracy of location by 10% Used Python (folium \u0026amp; Pandas) to build an interactive UI and map for rescue operators to track people in real time.","title":"Software Engineering Lead"}]