This research attempts to automate a portion of the design process for construction projects that follow Systems Engineering. Systems Engineering divides each stage into three main processes. They are the definition of requirements, functional analysis and allocation and design synthesis. This research attempts to automate the functional analysis and allocation because this process is repetitive, laborious, and time-consuming. The proposed system uses natural language processing to analyse the requirements and allocate them to objects in object type libraries (OTL). The OTL that is used in this research uses the OTL developed by Rijkswaterstaat. This OTL contains typical objects and sub-objects that an asset should include along with its metadata. This metadata constitutes a description of the object. The proposed system uses a dependency parser to add syntactic information to the requirements and object description. The requirements and objects are then subjected to Noun Chunking. Noun Chunking is an operation that breaks down the requirements and object description into meaningful phrases. Next, these chunks are compared for semantic similarity by plugging it into a word embedding to allocate each chunk a vector. The similarity is measured by finding the dot product of the two vectors. This system is converted into an NLP pipeline that identifies the correct requirement chunks to object chunk. The system is then compared to a neural classifier. Verification and validation are performed by measuring precision and recall, along with semi-structured interviews with industry experts. This research concludes that the proposed system can use process improvements. Still, the word embedding and cosine similarity is the suitable technology for the automating the allocation of requirements to object libraries.