Robotic missions to Mars and the Moon will increase over the next several decades. Rovers are a critical tool used in these missions, and they are constrained by challenges stemming from wheel slippage and slip-mitigation. This dissertation introduces three contributions to address these areas. First, I present a novel metric called \virtual slippage" to quantify wheel slippage by relating it to an effective change in wheel radius. Secondly, I study the feasibility of using auto and cross-correlation to detect wheel slippage. I found autocorrelation insufficient for slippage detection; I used cross-correlation and raw sensor data to train a neural network to detect slippage with an error of 0:07 to 0:1 classic slippage, or 0:02 to 0:5 cm using virtual slip on a 10 cm diameter wheel. Thirdly, I study the efficacy of asymmetrical wheel designs to mitigate and aid in escaping high-slip environments. By expanding and introducing the new defi nition of wheel slippage and new detection methods, this research paves the way for improved traction control on granular surfaces and improved maneuverability of extra-planetary rovers. I developed these tools to improve the success of future exploration missions for celestial bodies and improvements for terrestrial applications here on Earth.