Skip to main content

Data Science, Engineering and What’s in Between





I’m about to finish my first year in the data science program at the University of Nicosia (part-time) and my 6th year as a telecom engineer at Dar (full-time). During the past year, I have seen things that totally shifted my career focus and I want to share my top takes on the topic and the future. 


Top-down vs Bottom-up in corporations:


Applying data science/AI/ML at work as an engineer and a data scientist in the making is vastly more different from hiring a pure data scientist to help engineers with their work.


I see companies trying to catch the trend of ML/AI by hiring data scientists who don’t know anything about the engineering job and who start making “sophisticated” applications that really have minimal value and is mainly for show off.


I have been creating ML/SQL/deep learning tools at work with very minimal resources and even minimal coding because I know how the output should be, what is the data and how the job works, and how others can use these tools after me.


Some were  even shocked from 3 lines of SQL code I wrote the other day that automatically tracks changes in the building models across several stages. It’s as simple as:


Finding new/modified/deleted entries in tables which are the basics of SQL yet could save hours of work per project.


How to Kill Negativity with Data


Clash resolving is one of the most hectic and complex tasks in 3D design, ask any design engineer and he will confirm/cry.


My learnt lesson: when the problem is too complex: use data visualization.


It really gives a holistic view of the situation which ultimately reduces the stress caused by the few problems. By viewing the solved vs unsolved issues, we kill the our asymmetric habit of focusing on the negative stuff.


Below is a couple of plots I created from a recent project that was used to showcase, unlike what the leading department was claiming, that we were doing a good job in our Revit modeling: 





And the below plot highlighted the congested areas of clashes, which helped us focus on these areas to reduce 90% of the problems:





Use Assistants not Assassins 


Always remember: smart tools (AI/Ml/generative design) are meant to be assistive, not to replace people. 


Just like how Steve Jobs referred to computers as bicycle for the mind.


So, don’t fall for the fallacy that building tools should replace people, but you are doing it to help them be faster and more accurate, the human factor is very important to clip your risks.





Welcome to Mediocrestan


Our field of work is not an extremestan area. 


The other day I wanted to use data to see how much effort is being spent per project, and to my surprise, I saw that the distribution of Telecom/smart services elements within a building follow a normal distribution, which I think applies to MEP modeling in general.


Here is a plot of time required per location of elements:





(Using discretization for spatial data, time here is not sequential but rather fixed work efforts. So 10 seconds here means: 250 elements require approximately 10 second each to be added to the model.)


So, don’t be afraid of applying linear and some non-linear models to interpret data. We don’t have any outliers. And if there are outliers, they are anomalies.



Generative Design Conclusions.


My last take is on AI and the laptop class.


The beauty in fields like engineering design is the (relative) ease in collecting data from the tools used at work, like the ones developed by AutoDesk.


Computer programs run on tables. Any app that helps you export and use these tables, is opening up the opportunity to do wonders for the job.


For example, within a couple of hours, I was able to collect around 7K in room names from previous projects & whether they need to be secured or not. A simple preprocessing of data and a decision tree classifier achieved 91% accuracy in predicting whether a room needs to be secured or not.


But what about Neural Networks? They can do wonders as well. A simple model achieved around 92-93% accuracy with minimal efforts. And this is just the start.


So, as the laptop class was one of the classes who helped develop and advance AI, it is now affecting their jobs, adaptation is essential, by companies and individuals, this is a tail risk for everyone in this field to be aware of and prepare for.



I’m very excited for the next stages in interaction between BIM and data science. But remember, it won’t be data scientists who will lead the way, but rather engineers equipped with data science skills.


Peace/Eirini/Salam

Edmond