How Data Science Benefits MEP Engineers in Construction Industry

With every passing day, we keep realizing that data science has found a wide range of applications in numerous businesses. Seeing how the number of technological advancements is rapidly growing, the world has become overloaded with data. This is where data science shines. Construction industry has been a victim of poor planning, budgeting, miscalculations, proportional errors and low return on assets. Data science has the ability to overcome these issues and facilitate construction on every individual level.

The true value of data science lies in effectively processing a lot of information and extracting meaningful useful data from the lot. It becomes almost impossible for humans to process such large volumes of data with the speed of computers and extract useful data from raw data that has little to no use by themselves. Data science enables businesses with algorithms that helps in making effective business decisions. Also, data science coupled with computer graphics can help to understand the data easily with visuals.

One such promising data science application in MEP engineering is creating a Digital Twin of already existing buildings. In simple words, a digital twin in construction is a digital replica of an already existing building. A digital twin is considered to be a better and advanced version of a BIM model. BIM model is finalized by as-built aspect on completion of a project. However, a digital twin continuously updates itself by gathering measurements from the building sensors which provides a constant stream of data that can be analyzed to make buildings better.

Planning Building Upgrades by Using Data

The data collected from the various elements of the building can be used to predict possible future improvements and effect of modification like energy retrofits. This allows the building owner to virtually simulate a number of possible modifications and also observe their impact before even making an investment decision.

In old traditional projects, building upgrades were planned based on outdated documents and personnel visual inspection. Due to the complexity of buildings, it was highly probable that key information might be missed even after the inspection conducted by professionals. This would ultimately lead to change orders and unplanned costs due to lack of information during the building modifications.

During a building upgrade, there can be several upgrade options for the same building, and most of them being mutually exclusive. The following are 2 such examples:

• Heating systems can be based on use of only electricity or even a mix of electricity and combustion. Depending upon the local price for electricity, and availability of fuels and natural gases, the most economic option may vary.
• To deliver or remove heat, building’s HVAC systems can use different heat-transfer fluids like direct expansion systems use air ducts, hydronic systems use water piping, etc. The most suitable option may vary based on the conditions of the project.

Data science is capable of analyzing information and component behaviors that are invisible to humans. One such application is Energy Disaggregation. It is basically estimating the individual power consumption of each device by breaking down the electricity consumption if power meters. Energy disaggregation also allows virtual submetering which allows building owners to monitor energy consumption of every tenant. This information can help in identifying opportunities to save electricity.

Data Science in Building Design Process

MEP engineers can use data science for buildings that are still at as-design stage even before the construction starts. In such cases, design specifications are used for the building model instead of actual measured data. At such early stages, data science allows comparison of various design options that can be simulated for better understanding in a short amount of time. Although, such simulations are highly complex when there is no actual measured data, a digital twin of an already existing building can be used as a reference. This makes data science backed simulation a powerful tool for both planned buildings as well as existing buildings.

Data science can also assist with troubleshooting during building performance issues. The component data can be processed to find hidden interactions between the issues, allowing faster detection for implementing effective solutions.