971694Intelligent Environmental EstateClosedSmall Business Research InitiativeInnovate UKHighways England aspires to operate an intelligent, self-maintaining network of assets with a positive environmental footprint. How will automated systems deploy the right resources, to the right location at the right time? How will environmental value be quantified, improvement areas identified, prioritised and monitored? Understanding location and context is the foundation of intelligent operations. This project will test the feasibility of developing automated processes that can identify assets and understand the surrounding environment using multi-scale aerial and satellite imagery. There is a plethora of imagery products available and new services will allow Highways England to visualise environmental parameters in more detail, more often and more cost effectively. The proposed project will review current and future imagery products, evaluating their suitability for use in Highways England specific applications. This will identify potential environmental parameters and their available spatial (e.g. local or network wide) and temporal resolution. Using available imagery sources the project will develop machine learning algorithms, trained using the known locations of Highways England assets and environmental datasets. The project will build on Ramboll’s SiteSee platform which can analyse large volumes of remotely sensed data. It can identify and quantify ground cover type, vegetation classes, and infer aspects such as growth rates. The accuracy of the results will be verified using driven imagery and in-situ checks. The technical requirements of scaling the process across the entire network will also be evaluated. Using Ramboll domain knowledge and a deep understanding of Highways England's environmental obligations, the feasibility of using the results in a range of automated environmental applications will be tested. Applications include intelligent maintenance (e.g. condition of drainage infrastructure, monitoring invasive species treatment) and producing a regularly updated, consistent environmental baseline for improving biodiversity. In addition to applications providing a live view, the potential for evaluating future scenarios will be considered. One suggested priority is to automate the process of identifying areas where different types of planting achieve better environmental outcomes. Example metrics include noise, natural capital, carbon offset and biodiversity net gain. Automating these metrics would enable further innovation and revolutionise how Highways England manages its environmental estate.