This dissertation is aimed toward furthering biophysical studies using novel instrumentation techniques and algorithms. This work describes the development of optical tweezers as an instrumentation platform to study single bio-molecules. Algorithms for signal processing, data analysis are also developed in this thesis within the purview of single molecule experiments but also extensible to a wide array of applications. These tools are tested on experimental data obtained from samples prepared using protocols that are also a part of this work. Key contributions of the thesis includes the following: (i) Construction of optical tweezers with an emphasis on measures taken to reduce the unwanted extrinsic noise in the measurements. (ii) Development of an artificial neural network based technique to increase the measurement range of the instrument by twice the previously reported value. (iii) Application of a recursive least squares based approach to estimate the persistence length of a double stranded DNA molecule in real-time. (iv) Theoretically analyzing a coupled oscillator system as a sensor compared to a single oscillator system. (v) Development of analysis tools and experimental schemes to extract parameters that model kinesin flexibility. (vi) Development of a three-dimensional, multi-motor simulation environment to understand complicated dynamics of multi-motor transport and better interpretation of experimental data. (vii) Development and analysis of an algorithm to fit step signal to a noisy data from single molecule experiments that give better fitting than existing algorithms. The step detection algorithm is further extended to detect events like sudden changes in the parameters of the system in presence of non-linearities.