A new cooling system for a fleet of scientific instruments in the form of miniature wireless robots designed for interactions at the nanometer-scale is assessed to determine its limitations. Unlike other approaches, the use of a cooling chamber allows us to remove an embedded cooling system and maintain the overall size of each robot to a minimum, hence increasing the density of instruments per surface area and resulting in enhanced performance of the platform. The goal of this paper is to assess the capacity of this cooling system; not only to remove heat but also to reduce temperature fluctuations and difference in temperature levels among the robots to maintain each robot within an operational temperature range of 0-70degC. One hundred dummy robots were therefore placed in a custom-built cooling chamber which uses forced air convection. The temperature levels of the dummy robots were recorded with power dissipations from 0 to 15 W/robot and a maximum air flow rate of 0.5 m/s. It was determined that the maximum range in difference in temperature levels among the dummy robots increases by ~20degC per additional 5 W/robot of power dissipation with an initial difference of ~40degC at 5 W/robot. An estimated total power dissipation of 10 W/robot was determined to be a safe limit in order to maintain the operating temperature range of the robots between 0-70degC. For power dissipation over 10 W/robot, additional compensation methods are required. Although this study is concerned with the cooling of a fleet of miniature robots, the results and solutions proposed here could potentially be applied to other distributed systems where cooling is necessary. More specifically, for the cases where cooling of individual units is not practical nor possible due to higher costs and/or technical constraints, to name but a few reasons, this paper proposes a global convection cooling approach of all units performed within a chamber. The experimental data can be used to assess the viab- ility of this approach to other situations and help identifying potential issues prior to the implementation. Furthermore, the experimental data can be used to provide additional information on issues and challenges facing the optimal distribution of such units constrained to operate within relatively tight operating temperature ranges, and the limits of such approach including potential issues related to temperature fluctuations caused by local turbulences.