Master Thesis 30hp: Information Extraction during CAM Programming for the Generation of Expert Agents
One trend in aerospace is the machining of aircraft fuselages which saves time during the assembly. However, this requires time-intensive preparations due to the complexity of Computer-Aided Manufacturing (CAM). Since the number of machined parts increases, the CAM programming time also increases. This can be a critical bottleneck for the economic operation of the factory.
Together with KTH and Brogren Industries, Saab Aeronautics researches Expert Agents for the CAM programming environment within the R&D project IntDemo Flygplan. This project aims to support programmers by offering decision support and domain knowledge alongside their CAM programming tasks.
Previous work analyzed the requirements and potentials of Expert Agents workin alongside the CAM programming environment. A first software prototype is available for the observation of CAM programmers and tracking their workflows. This prototype uses screen filtering and Convolutional Neural Networks to record the programming actions and monitor the programmer’s workflows. All actions are labeled and recorded as image streams. Finally, based on the current programming state, a recommendation is generated from a predefined database.
Figure SEQ Figure \* ARABIC 1 Example interface to the CAM programming environment with highlighted regions of interest
This thesis builds atop the software prototype and tackles the challenges of further knowledge extraction. The successful candidate will firstly investigate Machine Learning algorithms for the optical character recognition (OCR) in the image stream of the identified and labeled programming actions. The main challenges are the automatic detection of relevant regions within the screen recordings and the link of identified Product Manufacturing Information (PMI) to the specific programming tasks. Additionally, the successful candidate will research different approaches for keeping track of the state of the CAM program. The thesis results will be integrated into an updated software prototype allowing for better recommendations
This master thesis work can start on short notice. Please don’t hesitate to send your application!
Contact at KTH: Bernd Peukert, email@example.com , tel. +46-793-41 43 47
Contact at Saab Aeronautics: Jesper Birberg, firstname.lastname@example.org , tel. +46-73-418 38 07
This position requires that you pass a security vetting based on the current regulations around/of security protection. For positions requiring security clearance additional obligations on citizenship may apply.
If you aspire to help create and innovate whilst developing yourself in a learning culture, Saab may well have the perfect conditions for you to grow. We pride ourselves on a nurturing environment, where everyone is different yet we share the same goal – to keep people and society safe.