A photogrammetry system for an accurate Indoor Positioning System for Field mapping experiments.

The scope of this project is a feasibility study for a photogrammetric system to be used in magnetic field mapping experiments. In the CERN Detector Technology group a team which is tasked with measuring magnetic fields for the magnets of different experiments. Examples of such magnets are solenoids, dipoles, toroids, etc. In order to measure the magnetic field, one makes use of a moving gantry. A beam is usually made to move on rails. On the beam a gantry is installed, which sweeps the volume of the magnet. On the gantry several magnetic field (hall) sensors are mounted. As the volume is swept, a map of the magnetic field is generated. Two information are needed for generating a field map. The first one is the magnetic field intensity, which is provided by the hall sensors, the second is the position (x, y, z) or (z, r, θ) depending from the magnet geometry. This position is provided by encoders mounted on the gantry. In order to have a good accuracy, the whole mechanical support needs to be machined with a very high precision. Small positioning errors might importantly affect the final magnetic field map. In this project we explore the possibility to drop the accuracy requirement on the mechanical structure, and make the sensor assembly "position aware". This belongs to the realm of "indoor positioning systems". The accuracy required is better than 1 mm. One important constraint is that all parts of the system need to be compatible with strong magnetic fields (> 1 T). This rules out most electric motors (with the only notable exception of piezoelectric motors). We explore here the idea to use photogrammetry in order to estimate the pose of the magnetic field probe. We assume that a target is placed somewhere at the end of your magnet (assuming for simplicity a solenoid topology, e.g. the solenoid of the CMS experiment). The volume is swept with magnetic field sensors, with a fixed focus camera mounted with the sensors. The camera is looking at a target at the end of the magnet, with some pattern on it, and is thus able to determine its position in space. Aim of this project is to quantitatively assess the resolution and accuracy limits for such a system.