During my internship at Geotab, I analyzed eMMC memory performance in embedded telematics devices by parsing diagnostic data uploaded to Google Cloud. The objective was to assess wear distribution across memory blocks after thousands of read/write cycles during long-term stress testing. I created a live-updating dashboard to visualize this data over several weeks, using Python and Colab to map erase counts and detect variability across memory regions. The analysis used read counts to evaluate randomness, wear leveling effectiveness, and identify potential bad blocks through cumulative distributions and histograms. This work supported failure risk detection and optimization of memory management in embedded systems.