Miniaturized transponder systems for mapping hydraulic fractures, monitoring unconventional reservoirs and measuring other wellbore parameters are under development. These devices, called FracBots (Fracture Robots), are envisioned as an extension of RFID (Radio Frequency IDentifcation) tags to realize Wireless Underground Sensor Networks (WUSNs) for mapping and characterization of hydraulic fractures in unconventional reservoirs. When injected during hydraulic fracturing operations, autonomous localization algorithms could be used to build up 3D constellation maps of proppant bed placement. To explore this concept, a FracBot prototype platform was developed, with which three key functions have been demonstrated, and are here reported. First, we developed a novel cross-layer communication framework for Magnetic Induction (MI) -based FracBot networks in dynamically changing underground environments, combining joint selection of modulation, channel coding and power control, a transmitter-based CDMA scheme and a geographic forwarding paradigm. Second, we developed a novel MI-based localization framework which exploits the unique properties of the MI field to determine the locations of the randomly deployed FracBot nodes. Third, we developed an accurate energy model framework for a linear FracBot network topology that gives feasible FracBots' transmission rates and FracBot network topology while respecting harvested energy constraints. Future work will include design and fabrication of miniaturized MI-based FracBot nodes for evaluation in a physical WUSN testbed.