TECHNOLOGY AREAS: Space Platforms; Air Platform; Ground Sea OBJECTIVE: This topic seeks to develop software that will allow an arbitrary number multiple mobile robotic manipulators to dynamically and intelligently team and cooperate for performing discrete tasks on a single asset. DESCRIPTION: Multi-Agent Simultaneous Segmentation and Scheduling (MASSS) is a complex problem in robotics, that if solved, will facilitate collaborative task execution by multiple mobile robots with shared or overlapping work envelopes to reduce the cycle time for low-volume, high-mix, and/or high-variability manufacturing processes. The current commercial, off the shelf (COTS) state-of-the-art for multi-robot systems is preprogrammed, pre-choreographed motion solutions that cannot accommodate the variable, uncertain, and stochastic processes encountered in defense aerospace manufacturing and sustainment operations. The motivation is the fact that cycle time is a cardinal performance metric in defense production, as it directly influences how quickly both the industrial and organic manufacturing bases can get materiel into the hands of the warfighter. Manufacturers and depots frequently deploy multiple human workers to collaboratively complete a given manufacturing process to reduce cycle time. A new class of agile, adaptive robots (e.g. scan-and-plan) is becoming available, but they are typically no faster, and often slower, than a single human worker. Without the ability to collaboratively work in parallel, the cycle time penalty they impose render applications that are otherwise ripe for robotic automaton as impractical. The central technical challenge associated with MASSS is to solve the simultaneous segmentation and scheduling (S3) problem on a time scale that enables intelligent dynamic task reallocation in response to process variation during system operation. S3 in an optimization problem that simultaneously considers: 1) the distribution of actions in both physical and task space amongst multiple agents; 2) the order in which agent should perform its assigned actions; and 3) system constraints, including but not limited to those associated with collisions, manipulator kinematics, manufacturing process, and precedence. The objective of this effort is to achieve the above goal for discrete manufacturing processes. Discrete processes are those that can be decomposed into a set of discrete robot poses and actions. Examples include de-fastening, hole making, and spot welding. In contrast, examples of continuous processes would include coating, de-coating, sanding, and blasting. The high-level project objectives are: 1) Demonstrate in simulation that a step-change in discrete process cycle time is feasible by dynamically reallocating discrete tasks between multiple robots operating in a shared workspace. 2) Develop reusable methods, tools, frameworks, etc. that can be used or adapted to a multitude of discrete manufacturing processes and system components. PHASE I: As this is a Direct-to-Phase-II (D2P2) topic, no Phase I awards will be made as a result of this topic. To qualify for this D2P2 topic, the Air Force expects the applicant(s) to demonstrate feasibility by means of a prior Phase I-type effort that does not constitute work undertaken as part of a prior or ongoing SBIR/STTR funding agreement. The feasibility study shall prove through prior research and development that the applicant has demonstrated mobile robotic manipulator technology with the following capabilities: (1) full 3 degree of freedom (DOF) base with a 6 DOF manipulator; (2) capability to accomplish aerospace-relevant discrete processing tasks; (3) localization, navigation, safety, and control systems needed for single-robot operation. Proof can be provided by direct demonstrations of any or all of the aforementioned capabilities, or through a combination of direct demonstrations and proposed modifications to an existing system to achieve the project goals. In the latter case, the applicant shall provide in-depth details on modifications necessary to achieve the desired capability. Modifications should utilize commercially available technologies. PHASE II: Eligibility for D2P2 is predicated on the applicant having performed a Phase I-like effort predominantly separate from the SBIR Programs. Under the phase II effort, the applicant shall sufficiently develop the technology in order to conduct a small number of relevant demonstrations in simulation. The final prototype system shall be capable of (1) Simultaneously optimizing task allocation and scheduling between multiple mobile robotic manipulators; (2) Supporting an arbitrary number of mobile manipulators; (3) Adapting to the stochastic nature of manufacturing processes (e.g., variability in individual task cycle time, replacing broken or worn tooling, variability in robot repositioning/motion time, etc.); and (4) Displaying simulation results graphically and predicting total cycle time for a given number of mobile manipulators. The awardee(s) must identify technology hurdles they are expected to encounter during the development program, as well as potential solutions to mitigate risk to the program. PHASE III DUAL USE APPLICATIONS: The awardee(s) will implement the MASSS software tools developed under the Phase II effort in hardware for multiple mobile manipulators and validate performance and achievement of objectives in pilot production at a USAF air logistics center. The contractor will pursue commercialization of the various technologies developed in Phase II for transitioning the technology to various defense aerospace OEMs, their supply chain, and the Air Force and broader DoD organic industrial base. Direct access with end users and government customers will be provided with opportunities to receive Phase III awards for providing the government additional research & development, or direct procurement of products and services developed in coordination with the program. REFERENCES: 1. Bui, H., Pierson, H.A., Nurre, S.G., and Sullivan, K.S. (2021) Toolpath Planning for Multi-Gantry Additive Manufacturing. IISE Transactions, 53(5), 552-567; 2. Bui, H., Pierson, H.A., Nurre, S.G, and Sullivan, K.M. (2020) Tool Path Planning Optimization for Multi-Tool Additive Manufacturing. Procedia Manufacturing, 39, 457-464; 3. Jin, Y., Pierson, H.A. and Liao, H. (2019) Toolpath allocation and scheduling for concurrent fused filament fabrication with multiple extruders. IISE Transactions, 51(2), 192-208. KEYWORDS: multi-robot collaboration; multi-robot optimization; multi-robot scheduling
