From an mechanical engineering standpoint, there are two interesting parts in the Experiment.
First of all, the sensor itself is a mechanical system, which ensures that the accelerations can be seen by the FBGs. Basically we have a fiber, containing the FBG, which is stretched between the sensor housing and a seismic mass. When the system accelerates, the inertia of the mass causes a force on the fiber. This force strains the fiber and the strain can be measured by the FBGs.
The other part is the module itself. It has to provide space and mounting points for the different subsystems, all while ensuring the survival of all parts during the strong accelerations of launch, the rising temperatures during ascend and descend, the dropping of air pressure to nearly vacuum and the shocks of parachute deployment and impact.
The biggest problems? Well, there is not that much space available in the module. We added a second floor so we were able to fit everything in. The other problem lies in the nature of the fibers itself. For example, the more you bend them, the more light is lost from the fibers core, decreasing the lights intensity and making it harder to measure. To connect fibers they have to be spliced. That means you have to carefully prepare the two ends of the fibers you need to join, ensuring that they are plain and clean, and then connect them by arc welding. Those joins are not that easy to make, sometimes it takes several tries, which can be a huge problem if you have an expensive part like the light source on the other end of the fiber. Also, every connection damps the light intensity further.
If you look at the picture above, you see the result of those problems. The fibers are routed through the module giving it a bit of a chaotic look – but it is ensured that the fibers aren’t bent to tight and the number of joins is minimized.