Understanding the complex interactions between atmosphere, snow, sea ice and ocean is one of the biggest challenges in polar research. These interactions control the observed changes in the Arctic and Antarctic sea ice cover, with consequences far beyond the Polar Oceans. But the lack of simultaneous in-situ observations leads to significant knowledge gaps on these interactions and their impacts. Here we present the concept and first results from our Multidisciplinary Ice-based Distributed Observatory (MIDO), which is a network of autonomous platforms that monitor the most essential climate and ecosystem parameters. A number of innovative instruments record atmosphere, snow, sea ice, and ocean parameters year round, including the largely under-sampled winter period. Data are publically available in near real time and contribute to numerical weather predictions and re-analysis data sets through the Global Telecommunication System. First platforms were installed on ice floes in the central Arctic Ocean in 2015. Their results suggest that this approach has great potential to advance our understanding of many physical and biogeochemical processes and interactions in the Polar Oceans. The time series indicate the timing and importance of different processes over the seasons along single drift paths. However, the ultimate aim is to achieve a quasi-synoptic, basin-wide coverage of key parameters, such as air temperature, barometric pressure, wind speed and direction, ice and snow thickness, incoming, reflected and transmitted irradiance, seawater temperature and salinity, Chl-a and CDOM fluorescence, turbidity, oxygen and nitrate. We will also discuss the challenge of maintaining a high data quality and present our plans to extend autonomous distributed observatories in the framework of the Year of Polar Prediction (YOPP, 2017-2019) and the Multidisciplinary drifting Observatory for the Study of the Arctic Climate (MOSAiC, 2019-2020).