The use of Type Ia supernovae as distance indicators led to the discovery of the accelerating expansion of the Universe a decade ago. While large 2nd generation surveys have significantly increased the size and quality of the high-redshift sample, the critical low-redshift reference sample useful for cosmology remains small. The Nearby Supernova Factory (SNfactory) is a US-France collaboration conducting a search for, and follow-up of, new nearby Type Ia supernovae in order to strengthen the statistical power from low-redshift supernovae. An equally important goal of the SNfactory is to develop better techniques to help control systematic uncertainties and better understand the underlying physics associated with Type Ia supernovae as distance indicators. To meet these goals the SNfactory has developed two new approaches to the problem. The first is an ultra-wide-field search which, like those at high redshift, accepts candidate supernovae regardless of any lack of association with a known host galaxy. The second is obtaining lightcurves from spectral time series for each supernova. This spectral time series approach side-steps some current limitations of existing supernova measurements based on broadband filters, and at the same time reveals valuable information on supernova physics from the time-evolution of the spectral features of each supernova. In this talk I will discuss this approach in more detail and show some recent initial results.