Picosatellites have only recently become a viable research topic thanks to the creation of the cubesat standard in 1999 and improvements in technology. However they are still limited in application because there are no high performance active actuators available in the market that can satisfy the mass/power/budget constraints of a picosatellite. Space and power are limited in these satellites which means that hardware redundancy is not very practical. If actuator faults occur, analytical redundancy techniques should be employed to determine if, where, and how the fault(s) occurred. This thesis focuses on enhancing picosatellite actuator tehnologies, as well as presenting an alogrithm for fault detection, isolation, and identification of ACS actuators.A CMG cluster design is proposed to demonstrate the feasibility of using CMGs in picosatellites to enhance their performance. The proposed CMG cluster design weighs less than 100g, occupies less than 25% of a cubesat's volume, and theoretically consumes less than 1.5W and 1W of peak and average power respectively. Furthermore, it is capable of providing sufficient torque and momentum storage for picosatellite attitude control in LEO. Next a novel adaptive Kalman filter algorithm is presented that can be implemented with the EKF and UKF for linear and non-linear systems respectively. The algorithm performs parameter estimation with sequential adaptive estimation and fading memory mechanisms that allow it to track changes in faulty parameters even in the presence of high levels of measurement noise. Furthermore, it is capable of tracking continuously varying and instantaneous changes in parameters. Numerical simulations are carried out to verify the performance of the propsed CMG cluster design as well as the fault diagnosis algorithm. The capabilities of the filter are further demonstrated via its application to a systems identification problem for a nonosatellite RW prototype being developed at SSDC group.