• The Journal of Korea Robotics Society
• /
• v.14 no.1
• /
• pp.40-49
• /
• 2019

Reinforcement learning has been applied to various problems in robotics. However, it was still hard to train complex robotic manipulation tasks since there is a few models which can be applicable to general tasks. Such general models require a lot of training episodes. In these reasons, deep neural networks which have shown to be good function approximators have not been actively used for robot manipulation task. Recently, some of these challenges are solved by a set of methods, such as Guided Policy Search, which guide or limit search directions while training of a deep neural network based policy model. These frameworks are already applied to a humanoid robot, PR2. However, in robotics, it is not trivial to adjust existing algorithms designed for one robot to another robot. In this paper, we present our implementation of Guided Policy Search to the robotic arms of the Baxter Research Robot. To meet the goals and needs of the project, we build on an existing implementation of Baxter Agent class for the Guided Policy Search algorithm code using the built-in Python interface. This work is expected to play an important role in popularizing robot manipulation reinforcement learning methods on cost-effective robot platforms.

• Journal of the Korean Regional Science Association
• /
• v.9 no.1
• /
• pp.65-78
• /
• 1993

In this study, the significant and enduring concentration of federal R&D spending in metro-scale clusters across the nation is treated as evidence of the operation of a distinct industrial infrastructure defined by the ability of R&D performers to attract external funding and pursue the sophisticated project work demanded. It follows, then, that the agglomerative potential of these R&D concentrations -- performers and their support infrastructures -- requires a search for economic impacts guided by a different stimulative effects attributable to federal R&D spending may be that substantial subnational economic impacts are routinely obscured and diluted by research designs that seek to discover impacts either at the level of nation-scale economic aggregates or on firms or specific industries organized spatially. Therefore, this study proceeds by seeking to link the locational clustering of federal contract R&D spending to more localized economic impacts. It tests a series of models(X-IV) designed to trace federal contract R&D spending flows to economic impacts registered at the level of metro-regional economies. By shifting the focus from funding sources to recipient types and then to sector-specific impacts, the patterns of consistent results become increasingly compelling. In general, these results indicated that federal R&D spending does indeed nurture the development of an important nation-spanning advanced industrial production and R&D infrastructure anchored primarily by two dozed or so metro-regions. However, dominated as it is by a strong defense-industrial orientation, federal contract R&D spending would appear to constitute a relatively inefficient national economic development policy, at least as registered on conventional indicators. Federal contract R&D destined for the support of nondefense/civilian(Model I), nonprofit(Model II), and educational/research(Mode III) R&D agendas is associated with substantially greater regional employment and income impacts than is R&D funding disbursed by the Department of Defense. While federal R&D support from DOD(Model I) and for-profit(Model II) and industrial performer(Model III) contract R&D agendas are associated with positive regional economic impacts, they are substantially smaller than those associated with performers operating outside the defense industrial base. Moreover, evidence that the large-business sector mediates a small business sector(Model VI) justifies closer scrutiny of the relative contribution to economic growth and development made by these two sectors, as well as of the primacy typically accorded employment change as a conventional economic performance indicator. Ultimately, those regions receiving federal R&D spending have experienced measurable employment and income gains as a result. However, whether or not those gains could be improved by changing the composition -- and therefore the primary missions -- of federal R&D spending cannot be decided by merely citing evidence of its economic impacts of the kind reported here. Rather, that decision turns on a prior public choice relating to the trade-offs deemed acceptable between conventional employment and income gains, the strength of a nation's industrial base not reflected in such indicators, and the reigning conception of what constitutes national security -- military might or a competitive civilian economy.