Subordinate robot

An example of a hierarchically structured robot collective is the system developed by Seib et al. (2011). In that robot collective consisting of two robots, the Apex robot performed the role of instructing a Subordinate robot to navigate an unknown environment, with the objective of cleaning the environment. The Subordinate robot was imbued with minimal intelligence and the robot collective relied on the Apex robot s abilities to develop a SLAM (simultaneous localization and mapping) (Durrant-White Bailey, 2006)... 

The hierarchically structured collective consists of two robot types the Apex robot and the Subordinate robot. Within the collective, there is one Apex robot and possibly multiple Subordinate robots. As an example, the illustrated robot collective in Figure 1 consists of three robots. The number of Subordinate robots may be increased or decreased according to mission needs. 

The Apex robot shoulders most of the overall computational load performed within the collective. Therefore, the Apex robot should be imbued with adequate computational power to deal with the burden of handling most of the collective s computational needs, consequently increasing its level of sophistication. However, the enhanced capabilitiesbroughtonby increased sophistication are often accompanied by an increased possibility for breakdowns (Landauer, 1995 Wickens Hollands, 2000). Yet, the comparatively higher level of computational prowess offered by the Apex robot means that Subordinate robots in the hierarchically structured collective can be computationally simpler. As a re suit, Subordinate robots can carry simpler and less powerful computers, possibly decreasing their size and development costs (Khoshnevis Bekey, 1998). 

The relationship between the Apex and Subordinate robots (as well as the human supervisor) is best explained through the description of a... 

Stage 2 Subordinate robots communication to Apex robot... 

Stage 4 Apex robot communication to Subordinate robots... 

Stage 3 Waypoints Allocation by Apex Robot. Following the transmission of the relevant information from Subordinate robots to the Apex robot, the Apex robot executes Stage 3. Stage 3 can be further decomposed into two steps. 

In the first step, the Apex robot updates its global wa5rpoints list by using the lists consisting ofvisited or skipped waypoints received from each Subordinate robot as well as the Apex s own list of visited or skipped wa5rpoints. In Figure 3, such waypoints are represented by shaded squares. Once... 

Stage A. Apex Robot Communication to Subordi-nate Robots. In the Stage 4, each Subordinate robot receives a new local waypoints list from the Apex robot. However, if a cluster has been assigned to the Apex robot, then wireless communication is not necessary. Stage 5 All Robots Communicate to the Human Supervisor. In the final stage, each robot transmits its newly allocated waypoints to the control station to indicate to the human supervisor that it has successfully coordinated. 

By comparing the proposed coordination techniques for both types of robot collective, it is possible to see that in order for a Subordinate robot in the hierarchically structured collective to take part in a coordination episode, it needs only to be aware of its own position and the way-points, which it has visited or skipped. It does not need to maintain any representation of the larger mission area nor even be aware of the existence of other collective members besides the Apex robot. This simplicity translates into a number of advantages. Firstly, it reduees the likelihood of computer glitches or bugs on the Subordinate robot. Secondly, it means that the necessary eomputation performed within the Subordinate robot can quite likely be achieved with small and inexpensive microprocessors (rather than larger... 

Also, as Subordinate robots in the hierarchically structured collective communicate only with their Apex robot. Inter-robot interactions are limited to only between Subordinates and the Apex. As such, the use of the hierarchical organizational structure allows for close coordination between robots while at the same time possibly facilitating reduced communication between robots (Fox, 1981). At the time of writing this chapter, the experimentto verify this claim is being conducted. 

It is foreseeable that there would be a limit to the munber of Subordinate robots that can be included under the jurisdiction of an Apex robot. Beyond that limit, the Apex robot would run out of bandwidth for communication with its subordinates. In addition, the plans for a large collective may be too complex for the Apex robot s limited computing power to cope with. 

If the size disparity between Apex and Subordinate Robots is big enough, then the concept of marsupialism can be applied to the collective such that the Apex not only performs cumbersome computation for the collective but also acts as a mothership to deploy the Subordinate robots. 

Figure 19.5. In the course of developing the robotic platform ARCoSyn for fully automated synthesis and purification of compound arrays, a sample-oriented concept has been realized that subordinates sample functionality, thus avoiding complex transport processes between spatially separated individual functionalities. The central component is an industrial robot, which - in respect of flexibility (the gripper changing system for several centralized functions), work space (optimal utilization of available surface area, no need for a translation axis), precision, and loading capacity (option of using modules for both miniaturization and upscaling) - is adapted to the requirements of the laboratory automation concept.

---