Advanced video analysis for goods flows in ports

Purpose and goal:

The project explores the possibilities of using AI image analysis to capture goods and traffic flows to and from the Port of Gothenburg. A proof-of-concept is carried out to evaluate the feasibility and potential of using AI on video streams for capturing goods and traffic flows before investing resources and time into a full-scale system demonstrator.

The project intends to carry out a strategic analysis of which technology is available on the market and how it is used, as well as which port implementations are available and what benefits these bring. Expected results and effects:

The project results in: 1) increased knowledge of movements in and out of the port, 2) benchmarking with respect to implementations of image analysis technology in a port environment, 3) strategic benefit analysis with regard to e.g. emissions, vehicle tracking, logistics, etc., 4) a plan with exact reading points for the subsequent full system demonstrator, and 5) approaches and rules for collecting goods and traffic data.

Approach and implementation:

The project begins by benchmarking existing implementations in parallel as problem and scope are defined. In the project´s problematization phase, the potential of the solution is assessed, where increased knowledge of obstacles and opportunities for full-scale demonstration is ascertained and how the project contributes to changing the intended system with respect to the system dimensions.

The project performs a proof-of-concept to test whether the idea of following up on goods and traffic flows to and from the port is feasible, with subsequent analysis of KPIs.