Why am I writing this?
I am writing this post, because I think this topic is interesting for people that are reading blog posts at intopreneur.com. It is especially interesting for people that like data, programming and modelling in relation to policy/strategy making. What I will describe is also interesting for start-ups that want to specialize in this area, because it is a real niche.
I am a Master of Science student at the Delft University of Technology, faculty Technology, Policy & Management. My master is called Engineering and Policy Analysis and focuses on Modelling & Simulation and Policy & Politics. What I will describe in this blog post are things that I have learned during the first year of my master.
A summary of the main idea
The building blocks are Data, Programming, Modelling & Simulation and Results presentation. You can see the building blocks in figure 1.
You start with a large amount of data, by means of programming you clean and prepare the data. The prepared data will be used as input for your mathematical model. You simulate your model and you get results. Now it is important that you present your results nicely, so that decision makers will take over your recommendations. With this process you can turn data into information and influence decision making.
A comprehensive explanation of the main idea
If you are still interested, then please keep on reading. Because in this part I will go more in depth about this topic. The total process is described in figure 2. I will walk you through the figure, so that you understand what is happening. The whole process, which I described above comes down to what is shown in figure 2. This is called “Data Driven Hybrid Modelling & Simulation”.
- System data –> In our world we are generating more and more data. This can be real time data or static data. The problem is that most of this data is not turned into information. With the process I describe here, the data becomes useful.
- Programming 1 –> The generated data is most likely messy and cannot be used directly for modelling. The data first need to be analysed, cleaned and prepared. This can be done by means of programming. In essence you create a new data set, which is called usable data. The analyst then can decide which of the usable data will be used as input for the modelling and simulation part.
- (Exploratory) Modelling & Simulation –> The usable data is prepared and can be used for modelling. However, what do I mean with modelling? With modelling, you recreate a part of a system. For example, the European freight infrastructure or the spread of infectious diseases. With simulation you simulate your model and you try to detect what might happen over time. In essence you can explore the future. The three main modelling branches are: System Dynamics (aggregated modelling), Discrete Event Simulation (dis-aggregated modelling) and Agent Based Modelling (dis-aggregated modelling). Until today these branches are operating individually. However, there is a need for Hybrid Modelling, in essence combining all these techniques. This is something that will emerge in the coming years.
- Programming 2 –> Now you have your results from the models you build. The output data need to be organised and stored in a database.
- Model data –> You went from system data to model data, or actually real information that can be used for decision making.
- Reporting –> This is maybe the most important part, because if you cannot report and present your results nicely to decision makers, they will not be convinced about your recommendations.
In the end your recommendations as policy/strategy analyst should enrich and improve the decision making process. Nowadays, managers and policy makers base a lot of decisions on qualitative reasoning, which is good up until a certain point. By adding the quantitative component, which I described in this blog post, it is likely to assume that decision making becomes better and this leads to sustainable strategies and policies.
Why have I told you this?
The process as described in figure 2 is not yet reality. There is still a lot of work required to make this happen. However, there is a strong need in the business field for informed decision making, based on quantitative methods. I think talented people, with good analytical skills can set-up a profitable start-up in this business area. If you are interested, you can contact me and I can bring you in contact with the department of Technology, Policy and Management from the Delft University of Technology.
Thank you for reading!