IP Library › Granted Patent US 12,450,534
Granted Patent B2
US 12,450,534 · App. 17/380,132 · Granted Oct 21, 2025

Heterogeneous graph attention networks for scalable multi-robot scheduling

Inventors: Matthew C. Gombolay (Atlanta, GA); Zheyuan Wang (Atlanta, GA)
Assignee: Georgia Tech Research Corporation
G06Q10/06313B25J9/1656G05B2219/32335G05B2219/39001
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,450,534
App. No.
17/380,132
Granted
Oct 21, 2025
Kind
B2
Abstract

An exemplary scheduler system and method are disclosed that can schedule a plurality of heterogenous robots to perform a set of tasks using heterogeneous graph attention network models. The exemplary scheduler system and method can outperform other work in multi-robot scheduling both in terms of schedule optimality and the total number of feasible schedules found and also in a scalable framework that can be trained via imitation-based Q-learning operations. The exemplary scheduler system and method can autonomously learn scheduling policies on multiple application domains.

Claims (65)

1. A computer-implemented method to efficiently coordinate multi-robot tasks performed by a plurality of heterogeneous robot equipment of different types, the method comprising:

collecting, by a hardware scheduler, at each of a plurality of schedule-able time steps, a list of available heterogeneous robot equipment into a set of available heterogeneous robot equipment; and

performing, by the hardware scheduler, a plurality of simulations to iteratively select each heterogeneous robot equipment from the set of available heterogeneous robot equipment and assign respective ones of the multi-robot tasks to each of the selected heterogeneous robot equipment using a Q-network, wherein each simulated assignment comprises:

receiving, by the hardware scheduler, temporal constraints comprising: deadlines and wait constraints of each of the multi-robot tasks and an objection for a minimizing of an overall process duration of the multi-robot tasks;

receiving, by the hardware scheduler, spatial constraints comprising: a first spatial constraint requiring that a location can only be occupied by one of the heterogeneous robot equipment at a time and a second spatial constraint requiring a minimum safety distance allowed between each of the plurality of heterogeneous robot equipment while executing the multi-robot tasks;

inputting, by the hardware scheduler, into a simple temporal network (STN)-based model, a plurality of nodes corresponding to the multi-robot tasks, each represented by a start time node and a finish time node;

reducing, by the hardware scheduler, complexity of the simple temporal network (STN)-based model by removing all finish time nodes of the multi-robot tasks, except for a time point when all the multi-robot tasks will be completed while preserving all the temporal constraints or providing a reduced STN model of the STN;

building, by the hardware scheduler, a heterogeneous graph g from states in the STN-based model that convolutionality encodes into the heterogeneous graph g: the temporal constraints, the spatial constraints, and at least one other constraint associated with the available heterogeneous robot equipment, locations of the available heterogeneous robot equipment, specific locations of the multi-robot tasks, and shared tools employed by the available heterogeneous robot equipment;

computing, by the hardware scheduler using heterogeneous graph attention layers of a graph attention network, input features for the plurality of nodes in the heterogeneous graph g, wherein the input features comprises: a minimum of an expected time to complete an unscheduled task of a plurality of unscheduled tasks, a maximum of the expected time to complete the unscheduled task, a mean of the expected time to complete the unscheduled task, and a standard deviation of the expected time to complete the unscheduled task;

merging, by the hardware scheduler, the inputted features as a multi-head output for each multi-head layer in the heterogeneous graph attention layers of the graph attention network using at least a concatenation and an averaging;

learning, by the hardware scheduler, a greedy policy for sequential decision making by constructing a schedule as a Markov decision process (MDP) using a tuple that includes at least: a first tuple comprising the states at each decision-step that includes the temporal constraints represented by the STN, the locations of the available heterogeneous robot equipment, and all, previously constructed, partial schedules of the available heterogeneous robot equipment, a second tuple comprising actions corresponding to appending the unscheduled task at end of a partial schedule of a selected one of the available heterogeneous robot equipment, a third tuple comprising transitions, corresponding to adding edges associated with each of the actions into the STN and updating the partial schedule of the selected one of the available heterogeneous robot equipment, a fourth tuple comprising a reward of a state-action pair defined as a change in an objective value after taking one of the actions while minimizing the overall process duration and a fifth tuple comprising a discount factor;

employing, by the hardware scheduler, imitation learning to train the Q-network, by scaling up, the heterogeneous graph g from small-scale solutions employed by an expert, to large scale problems solved by grounding, below a total value of the reward, Q-values of alternative actions not selected by the expert while ensuring that a gradient is trained only on alternative, unselected actions with a maximum Q-value, and ensuring that the gradient is propagated through all the alternative unselected actions that have a subset of the Q-values higher than a difference between the total value of the reward and an empirically selected offset constant;

employing, by the hardware scheduler, the graph attention network, the input features, the transitions, and a result of the imitation learning to generate a greedy schedule; and

iteratively selecting, by the hardware scheduler, based on the greedy policy, the greedy schedule having assignments of the multi-robot tasks to each of the selected heterogeneous robot equipment, the greedy schedule in accordance with a policy from the group consisting of: a first policy associated with first availability of a first one of the available heterogeneous robot equipment, a second policy associated with a minimum average time on the plurality of unscheduled tasks, a third policy associated with a minimum time on any one of the plurality of unscheduled task, a fourth policy associated with a minimum average time on all of the multi-robot tasks.

2. The method of claim 1 , wherein the STN-based model comprises (i) a first graph portion that encodes the temporal constraints and (ii) second network portion that encodes the task and the available heterogeneous robot equipment.

3. The method of claim 1 , wherein the STN-based model encodes the temporal constraints and at least one constraint associated with spatial constraints, available robots, robot locations, task locations, and shared resources.

4. The method of claim 1 , wherein the STN-based model is built by:

generating a first graph comprising a plurality of task nodes comprising a start time node and a finish time node; and

generating a second graph as the STN-based model by removing the finish time node.

5. The method of claim 1 , wherein the simple temporal network STN-based model is built by:

generating a base graph comprising a minimum distance graph;

adding a plurality of robot nodes to the base graph, wherein each robot node of the plurality of robot nodes is connected to an assigned task node, and wherein each robot node of the plurality of robot nodes is connected to other robot nodes of the plurality of robot nodes;

adding a plurality of location nodes to the base graph, wherein each location node of the plurality of location nodes is connected to an assigned task node, and wherein each location node of the plurality of location nodes is connected to other location nodes of the plurality of location nodes; and

adding a plurality of state summary nodes to the base graph, wherein each state summary node of the plurality of state summary nodes is connected to a task node, a robot node, and a location node.

6. The method of claim 5 , further comprising:

adding a plurality of Q-value nodes to the base graph, where each of the Q-value nodes of the plurality of Q-value nodes is connected to a task node, a robot node, and a location node.

7. The method of claim 1 , wherein the STN-based model is generated in part using Johnson algorithm or Floyd Warshall algorithm to generate a minimum distance graph as a structure for the STN-based model.

8. The method of claim 1 , wherein the plurality of heterogeneous robot equipment comprise at least one of robotic equipment, manufacturing equipment, and transport equipment.

9. The method of claim 1 , wherein the plurality of heterogeneous robot equipment comprise machines in combination with one or more human workers with assigned tasks in manufacturing, assembling, distributing workflow.

10. A hardware scheduler system comprising:

a processor; and

a memory operatively coupled to the processor, the memory having instructions stored therein, wherein execution of the instructions by the processor causes the processor to:

efficiently coordinate multi-robot tasks performed by a plurality of heterogeneous robot equipment of different types, wherein the coordinating includes:

at each of a plurality of schedule-able time steps, a hardware scheduler collecting a list of available heterogeneous robot equipment into a set of available heterogeneous robot equipment; and

performing, by the hardware scheduler, a plurality of simulations to iteratively select each heterogeneous robot equipment from the set of available heterogeneous robot equipment and assign respective ones of the multi-robot tasks to each of the selected heterogeneous robot equipment using a Q-network, wherein each simulated assignment comprises:

receiving, by the hardware scheduler, temporal constraints comprising:

deadlines and wait constraints of each of the multi-robot tasks and an objection for a minimizing of an overall process duration of the multi-robot tasks;

receiving, by the hardware scheduler, spatial constraints comprising: a first spatial constraint requiring that a location can only be occupied by one of the heterogeneous robot equipment at a time and a second spatial constraint requiring a minimum safety distance allowed between each of the plurality of heterogeneous robot equipment while executing the multi-robot tasks;

inputting, by the hardware scheduler, into a simple temporal network (STN)-based model, a plurality of nodes corresponding to the multi-robot tasks, each represented by a start time node and a finish time node;

reducing, by the hardware scheduler, complexity of the STN-based model by removing all finish time nodes of the multi-robot tasks, except for a time point when all the multi-robot tasks will be completed, while preserving all the temporal constraints or providing a reduced STN model of the STN;

building, by the hardware scheduler, a heterogeneous graph g from states in the STN-based model that convolutionality encodes into the heterogeneous graph g: the temporal constraints, the spatial constraints, and at least one other constraint associated with the available heterogeneous robot equipment, locations of the available heterogeneous robot equipment, specific locations of the multi-robot tasks, and shared tools employed by the available heterogeneous robot equipment;

computing, by the hardware scheduler using heterogeneous graph attention layers of a graph attention network, input features for the plurality of nodes in the heterogeneous graph g, wherein the input features comprise: a minimum of an expected time to complete an unscheduled task of a plurality of unscheduled tasks, a maximum of the expected time to complete the unscheduled task, a mean of the expected time to complete the unscheduled task, and a standard deviation of the expected time to complete the unscheduled task;

merging, by the hardware scheduler, the inputted features as a multi-head output for each multi-head layer in the heterogeneous graph attention layers of the graph attention network using one of a concatenation and an averaging;

learning, by the hardware scheduler, a greedy policy for sequential decision making by constructing a schedule as a Markov decision process (MDP) using a tuple that includes at least: a first tuple comprising the states at each decision-step that includes the temporal constraints represented by the STN, the locations of the available heterogeneous robot equipment, and all, previously constructed, partial schedules of the available heterogeneous robot equipment, a second tuple comprising actions corresponding to appending the unscheduled task at end of a partial schedule of a selected one of the available heterogeneous robot equipment, a third tuple comprising transitions, corresponding to adding edges associated with each of the actions into the STN and updating the partial schedule of the selected one of the available heterogeneous robot equipment, a fourth tuple comprising a reward of a state-action pair defined as a change in an objective value after taking one of the actions while minimizing the overall process duration and a fifth tuple comprising a discount factor;

employing, by the hardware scheduler, imitation learning to train the Q-network, by scaling up, the heterogeneous graph g from small-scale solutions employed by an expert, to large scale problems solved by grounding, below a total value of the reward, Q-values of alternative actions not selected by the expert while ensuring that a gradient is trained only on alternative, unselected actions with a maximum Q-value, and ensuring that the gradient is propagated through all the alternative unselected actions that have a subset of the Q-values higher than a difference between the total value of the reward and an empirically selected offset constant;

employing, by the hardware scheduler, the graph attention network, the input features, the transitions, and a result of the imitation learning to generate a greedy schedule; and

iteratively selecting, by the hardware scheduler, based on the greedy policy, the greedy schedule having assignments of the multi-robot tasks to each of the selected heterogeneous robot equipment, the greedy schedule in accordance with a policy from the group consisting of: a first policy associated with first availability of a first one of the available heterogeneous robot equipment, a second policy associated with a minimum average time on the plurality of unscheduled tasks, a third policy associated with a minimum time on any one of the plurality of unscheduled task, a fourth policy associated with a minimum average time on all of the multi-robot tasks.

11. The hardware scheduler system of claim 10 , wherein the STN-based model comprises (i) a first graph portion that encodes the temporal constraints and (ii) a second graph portion that encodes the task and the available heterogeneous robot equipment.

12. A non-transitory computer-readable medium having instructions stored thereon, wherein the instructions, when executed by a processor, cause the processor to:

efficiently coordinate multi-robot tasks performed by a plurality of heterogeneous robot equipment of different types, wherein the coordinating includes:

at each of a plurality of schedule-able time steps, a hardware scheduler collecting a list of available heterogeneous robot equipment into a set of available heterogeneous robot equipment; and

performing, by the hardware scheduler, a plurality of simulations to iteratively select each heterogeneous robot equipment from the set of available heterogeneous robot equipment and assign respective ones of the multi-robot tasks to each of the selected heterogeneous robot equipment using a Q-network, wherein each simulated assignment comprises:

receiving, by the hardware scheduler, temporal constraints comprising:

deadlines and wait constraints of each of the multi-robot tasks and an objection for a minimizing of an overall process duration of the multi-robot tasks;

receiving, by the hardware scheduler, spatial constraints comprising: a first spatial constraint requiring that a location can only be occupied by one of the heterogeneous robot equipment at a time and a second spatial constraint requiring a minimum safety distance allowed between each of the plurality of heterogeneous robot equipment while executing the multi-robot tasks;

inputting, by the hardware scheduler, into a simple temporal network (STN)-based model, a plurality of nodes corresponding to the multi-robot tasks, each represented by a start time node and a finish time node;

reducing, by the hardware scheduler, complexity of the STN-based model by removing all finish time nodes of the multi-robot tasks, except for a time point when all the multi-robot tasks will be completed, while preserving all the temporal constraints or providing a reduced STN model of the STN;

building, by the hardware scheduler, a heterogeneous graph g from states in the STN-based model that convolutionality encodes into the heterogeneous graph g: the temporal constraints, the spatial constraints, and at least one other constraint associated with the available heterogeneous robot equipment, locations of the available heterogeneous robot equipment, specific locations of the multi-robot tasks, and shared tools employed by the available heterogeneous robot equipment;

computing, by the hardware scheduler using heterogeneous graph attention layers of a graph attention network, input features for the plurality of nodes in the heterogeneous graph g, wherein the input features comprise: a minimum of an expected time to complete an unscheduled task of a plurality of unscheduled tasks, a maximum of the expected time to complete the unscheduled task, a mean of the expected time to complete the unscheduled task, and a standard deviation of the expected time to complete the unscheduled task;

merging, by the hardware scheduler, the inputted features as a multi-head output for each multi-head layer in the heterogeneous graph attention layers of the graph attention network using a concatenation and an averaging;

learning, by the hardware scheduler, a greedy policy for sequential decision making by constructing a schedule as a Markov decision process (MDP) using a tuple that includes at least: a first tuple comprising the states at each decision-step that includes the temporal constraints represented by the STN, the locations of the available heterogeneous robot equipment, and all, previously constructed, partial schedules of the available heterogeneous robot equipment, a second tuple comprising actions corresponding to appending the unscheduled task at end of a partial schedule of a selected one of the available heterogeneous robot equipment, a third tuple comprising transitions, corresponding to adding edges associated with each of the actions into the STN and updating the partial schedule of the selected one of the available heterogeneous robot equipment, a fourth tuple comprising a reward of a state-action pair defined as a change in an objective value after taking one of the actions while minimizing the overall process duration and a fifth tuple comprising a discount factor;

employing, by the hardware scheduler, imitation learning to train the Q-network, by scaling up, the heterogeneous graph g from small-scale solutions employed by an expert, to large scale problems solved by grounding, below a total value of the reward, Q-values of alternative actions not selected by the expert while ensuring that a gradient is trained only on alternative, unselected actions with a maximum Q-value, and ensuring that the gradient is propagated through all the alternative unselected actions that have a subset of the Q-values higher than a difference between the total value of the reward and an empirically selected offset constant;

employing, by the hardware scheduler, the graph attention network, the input features, the transitions, and a result of the imitation learning to generate a greedy schedule; and

iteratively selecting, by the hardware scheduler, based on the greedy policy, the greedy schedule having assignments of the multi-robot tasks to each of the selected heterogeneous robot equipment, the greedy schedule in accordance with a policy from the group consisting of: a first policy associated with first availability of a first one of the available heterogeneous robot equipment, a second policy associated with a minimum average time on the plurality of unscheduled tasks, a third policy associated with a minimum time on any one of the plurality of unscheduled task, a fourth policy associated with a minimum average time on all of the multi-robot tasks.

13. The non-transitory computer-readable medium of claim 12 , wherein the STN-based model comprises (i) a first graph portion that encodes the temporal constraints and (ii) a second graph portion that encodes the task and the available heterogeneous robot equipment.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 6, 2021
From: GOMBOLAY, MATTHEW C.; WANG, ZHEYUAN
To: GEORGIA TECH RESEARCH CORPORATION
Reel/Frame 057712/0494 →
Continuity (3)
Provisional Application 63053958 · Jul 20, 2020
Provisional Application 63053954 · Jul 20, 2020
Related Publication 20220226994A1 · Jul 21, 2022
References Cited (125)
US 5666518A · Jumper · 1997 [cited by examiner]
US 9304817B2 · Kim · 2016 [cited by examiner]
US 9454157B1 · Hafeez · 2016 [cited by examiner]
US 9569736B1 · Ghesu · 2017 [cited by examiner]
US 9821455B1 · Bareddy · 2017 [cited by examiner]
US 10282864B1 · Kim · 2019 [cited by examiner]
US 11079245B1 · Niewiadomski · 2021 [cited by examiner]
US 20020118207A1 · Jagla · 2002 [cited by examiner]
US 20050188376A1 · Matsumoto · 2005 [cited by examiner]
US 20090265299A1 · Hadad · 2009 [cited by examiner]
US 20100094459A1 · Cho · 2010 [cited by examiner]
US 20100222924A1 · Gienger · 2010 [cited by examiner]
US 20100262574A1 · Zhou · 2010 [cited by examiner]
US 20140088763A1 · Hazan · 2014 [cited by examiner]
US 20140207282A1 · Angle · 2014 [cited by examiner]
US 20160232445A1 · Srinivasan · 2016 [cited by examiner]
US 20170076201A1 · van Hasselt · 2017 [cited by examiner]
US 20170154261A1 · Sunehag · 2017 [cited by examiner]
US 20180101957A1 · Talathi · 2018 [cited by examiner]
US 20180165603A1 · Van Seijen · 2018 [cited by examiner]
US 20180173563A1 · Hu · 2018 [cited by examiner]
US 20180253837A1 · Ghesu · 2018 [cited by examiner]
US 20190094532A1 · Jabbour · 2019 [cited by examiner]
US 20190230046A1 · Djukic · 2019 [cited by examiner]
US 20190392001A1 · Carothers · 2019 [cited by examiner]
US 20200003886A1 · Cho · 2020 [cited by examiner]
US 20200012382A1 · Lee · 2020 [cited by examiner]
US 20200027006A1 · Gupta · 2020 [cited by examiner]
US 20200104714A1 · Lee · 2020 [cited by examiner]
US 20200301013A1 · Banerjee · 2020 [cited by examiner]
US 20200303266A1 · Jeong · 2020 [cited by examiner]
US 20210012767A1 · Kupryjanow · 2021 [cited by examiner]
US 20210065440A1 · Sunkavalli · 2021 [cited by examiner]
US 20210073912A1 · Da Silva · 2021 [cited by examiner]
US 20210103286A1 · Wang · 2021 [cited by examiner]
US 20210110262A1 · Schmitt · 2021 [cited by examiner]
US 20210124344A1 · Hu · 2021 [cited by examiner]
US 20210271253A1 · Liu · 2021 [cited by examiner]
US 20210306005A1 · Sheiman · 2021 [cited by examiner]
US 20210365782A1 · Huang · 2021 [cited by examiner]
US 20210383176A1 · Wu · 2021 [cited by examiner]
US 20210383534A1 · Tadross · 2021 [cited by examiner]
US 20210398014A1 · Cao · 2021 [cited by examiner]
US 20220067640A1 · George · 2022 [cited by examiner]
US 20220164659A1 · Song · 2022 [cited by examiner]
US 20220188632A1 · Culurciello · 2022 [cited by examiner]
US 20220269254A1 · Putman · 2022 [cited by examiner]
US 20220355051A1 · Srinivasan · 2022 [cited by examiner]
US 20220390256A1 · Pan · 2022 [cited by examiner]
US 20220398921A1 · Jaggi · 2022 [cited by examiner]
US 20220412275A1 · Waldhart · 2022 [cited by examiner]
US 20230085147A1 · Melnikov · 2023 [cited by examiner]
US 20230182296A1 · Sermanet · 2023 [cited by examiner]
US 20230326599A1 · Togawa · 2023 [cited by examiner]
WO WO2020130909A1 · 2020 [cited by examiner]
Wang et al Learning to dynamically coordinate multi-robot teams in graph attention networks arXiv 1912.0205, Dec. 4, 2019 https://arxiv.org/abs/1912.02059 (Year: 2019). [cited by examiner]
Wang et al, Heterogeneous graph attention network. In The world wide web conference pp. 2022-2032, May 13, 2019 https://dl.acm.org/doi/abs/10.1145/3308558.3313562 (Year: 2019). [cited by examiner]
Zhao et al, Deep reinforcement learning for user association and resource allocation in heterogeneous cellular networks, IEEE Transactions on Wireless Communications 18, No. 11, pp. 5141-5152, Aug. 13, 2019 https://ieee… [cited by examiner]
C. Heyer, “Human-robot interaction and future industrial robotics applications,” in 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems. IEEE, 2010, pp. 4749-4754. [cited by applicant]
H. Raghavan, O. Madani, and R. Jones, “Active learning with feedback on features and instances,” Journal of Machine Learning Research, vol. 7, No. Aug, pp. 1655-1686, 2006. [cited by applicant]
E. Khalil, H. Dai, Y. Zhang, B. Dilkina, and L. Song, “Learning combinatorial optimization algorithms over graphs,” in Advances in Neural Information Processing Systems, 2017, pp. 6348-6358. [cited by applicant]
W. Kool, H. van Hoof, and M. Welling, “Attention, learn to solve routing problems!” in International Conference on Learning Representations, 2019. [cited by applicant]
P. Velickovic, G. Cucurull, A. Casanova, A. Romero, P. Lio, and Y. Bengio, “Graph Attention Networks,” International Conference on Learning Representations, 2018. [cited by applicant]
R. Dechter, I. Meiri, and J. Pearl, “Temporal constraint networks,” Artificial intelligence, vol. 49, No. 1-3, pp. 61-95, 1991. [cited by applicant]
E. Nunes, M. Manner, H. Mitiche, and M. Gini, “A taxonomy for task allocation problems with temporal and ordering constraints,” Robotics and Autonomous Systems, vol. 90, pp. 55-70, 2017. [cited by applicant]
G. A. Korsah, A. Stentz, and M. B. Dias, “A comprehensive taxonomy for multi-robot task allocation,” The International Journal of Robotics Research, vol. 32, No. 12, pp. 1495-1512, 2013. [cited by applicant]
P. Brucker, A. Drexl, R. M{umlaut over ( )}ohring, K. Neumann, and E. Pesch, “Resourceconstrained project scheduling: Notation, classification, models, and methods,” European journal of operational research, vol. 112, N… [cited by applicant]
J. F. Benders, “Partitioning procedures for solving mixed-variables programming problems,” Numerische mathematik, vol. 4, No. 1, pp. 238-252, 1962. [cited by applicant]
J. Chen and R. G. Askin, “Project selection, scheduling and resource allocation with time dependent returns,” European Journal of Operational Research, vol. 193, No. 1, pp. 23-34, 2009. [cited by applicant]
W. Tan and B. Khoshnevis, “A linearized polynomial mixed integer programming model for the integration of process planning and scheduling,” Journal of Intelligent Manufacturing, vol. 15, No. 5, pp. 593-605, 2004. [cited by applicant]
A. Kushleyev, D. Mellinger, C. Powers, and V. Kumar, “Towards a swarm of agile micro quadrotors,” Autonomous Robots, vol. 35, No. 4, pp. 287-300, 2013. [cited by applicant]
S. M. Mousavi and R. Tavakkoli-Moghaddam, “A hybrid simulated annealing algorithm for location and routing scheduling problems with cross-docking in the supply chain,” Journal of Manufacturing Systems, vol. 32, No. 2, p… [cited by applicant]
L. Zhang and T. Wong, “An object-coding genetic algorithm for integrated process planning and scheduling,” European Journal of Operational Research, vol. 244, No. 2, pp. 434-444, 2015. [cited by applicant]
W. Zhang and T. G. Dietterich, “A reinforcement learning approach to jobshop scheduling,” in Proceedings of the International Joint Conference on Artificial Intelligence, 1995, pp. 1114-1120. [cited by applicant]
Y.-C. Wang and J. M. Usher, “Application of reinforcement learning for agent-based production scheduling,” Engineering Applications of Artificial Intelligence, vol. 18, No. 1, pp. 73-82, 2005. [cited by applicant]
J. Wu, X. Xu, P. Zhang, and C. Liu, “A novel multi-agent reinforcement learning approach for job scheduling in grid computing,” Future Generation Computer Systems, vol. 27, No. 5, pp. 430-439, 2011. [cited by applicant]
H. Mao, M. Schwarzkopf, S. B. Venkatakrishnan, Z. Meng, and M. Iizadeh, “Learning scheduling algorithms for data processing clusters,” in Proceedings of the ACM Special Interest Group on Data Communication, 2019, pp. 27… [cited by applicant]
M. Gasse, D. Chetelat, N. Ferroni, L. Charlin, and A. Lodi, “Exact combinatorial optimization with graph convolutional neural networks,” in Advances in Neural Information Processing Systems, 2019, pp. 15 554-15 566. [cited by applicant]
J. Zhou, G. Cui, Z. Zhang, C. Yang, Z. Liu, and M. Sun, “Graph neural networks: A review of methods and applications,” arXiv preprint arXiv:1812.08434, 2018. [cited by applicant]
L. Barbulescu, Z. B. Rubinstein, S. F. Smith, and T. L. Zimmerman, “Distributed coordination of mobile agent teams: the advantage of planning ahead,” in Proceedings of the 9th International Conference on Autonomous Agen… [cited by applicant]
B. Coltin and M. Veloso, “Online pickup and delivery planning with transfers for mobile robots,” in Workshops at the Twenty-Seventh AAAI Conference on Artificial Intelligence, 2013. [cited by applicant]
E. Nunes and M. Gini, “Multi-robot auctions for allocation of tasks with temporal constraints,” in Twenty-Ninth AAAI Conference on Artificial Intelligence, 2015. [cited by applicant]
M. C. Gombolay, R. J. Wilcox, and J. A. Shah, “Fast scheduling of robot teams performing tasks with temporospatial constraints,” IEEE Transactions on Robotics, vol. 34, No. 1, pp. 220-239, 2018. [cited by applicant]
I. Tsamardinos and M. E. Pollack, “Efficient solution techniques for disjunctive temporal reasoning problems,” Artificial Intelligence, vol. 151, No. 1-2, pp. 43-89, 2003. [cited by applicant]
C. H. Papadimitriou and K. Steiglitz, Combinatorial optimization: algorithms and complexity. Courier Corporation, 1998. [cited by applicant]
B. Piot, M. Geist, and O. Pietquin, “Boosted bellman residual minimization handling expert demonstrations,” in Joint European Conference on Machine Learning and Knowledge Discovery in Databases. Springer, 2014, pp. 549-… [cited by applicant]
A. Paszke, S. Gross, S. Chintala, G. Chanan, E. Yang, Z. DeVito, Z. Lin, A. Desmaison, L. Antiga, and A. Lerer, “Automatic differentiation in PyTorch,” in NIPS Autodiff Workshop, 2017. [cited by applicant]
D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” arXiv preprint arXiv:1412.6980, 2014. [cited by applicant]
H. Hellerman, “Some principles of time-sharing scheduler strategies,” IBM Systems Journal, vol. 8, No. 2, pp. 94-117, 1969. [cited by applicant]
S. Wilson, P. Glotfelter, L. Wang, S. Mayya, G. Notomista, M. Mote, and M. Egerstedt, “The robotarium: Globally impactful opportunities, challenges, and lessons learned in remote-access, distributed control of multirobo… [cited by applicant]
Wang, Zheyuan, and Matthew Gombolay. “Learning scheduling policies for multi-robot coordination with graph attention networks.” IEEE Robotics and Automation Letters 5, No. 3 (Jun. 2020): 4509-4516. [cited by applicant]
Elkin Castro and Sanja Petrovic. Combined mathematical programming and heuristics for a radiotherapy pretreatment scheduling problem. Journal of Scheduling, 15 (3):333-346, 2012. [cited by applicant]
Imen Essafi, Yazid Mati, and St´ephane Dauzere-Peres. A genetic local search algorithm for minimizing total weighted tardiness in the job-shop scheduling problem. Computers & Operations Research, 35(8):2599-2616, 2008. [cited by applicant]
Robert W Floyd. Algorithm 97: shortest path. Communications of the ACM, 5(6):345, 1962. [cited by applicant]
Eduardo Feo Flushing, Luca M Gambardella, and Gianni A Di Caro. Simultaneous task allocation, data routing, and transmission scheduling in mobile multirobot teams. In 2017 IEEE/RSJ International Conference on Intelligen… [cited by applicant]
Yann LeCun, Yoshua Bengio, and Geoffrey Hinton. Deep learning. nature, 521(7553):436-444, 2015. [cited by applicant]
Adam Paszke, et al., PyTorch: An Imperative Style, High-Performance Deep Learning Library. In Advances in Neural Information Processing Systems 32, pp. 8024-8035, 2019. [cited by applicant]
Huizhi Ren and Lixin Tang. An improved hybrid milp/cp algorithm framework for the job-shop scheduling. In 2009 IEEE International Conference on Automation and Logistics, pp. 890-894. IEEE, 2009. [cited by applicant]
Marius M Solomon. On the worst-case performance of some heuristics for the vehicle routing and scheduling problem with time window constraints. Networks, 16(2): 161-174, 1986. [cited by applicant]
Minjie Wang, Lingfan Yu, Da Zheng, Quan Gan, Yu Gai, Zihao Ye, Mufei Li, Jinjing Zhou, Qi Huang, Chao Ma, Ziyue Huang, Qipeng Guo, Hao Zhang, Haibin Lin, Junbo Zhao, Jinyang Li, Alexander J Smola, and Zheng Zhang. Deep … [cited by applicant]
Xiao Wang, Houye Ji, Chuan Shi, Bai Wang, Yanfang Ye, Peng Cui, and Philip S Yu. Heterogeneous graph attention network. In The World Wide Web Conference, pp. 2022-2032. ACM, 2019. [cited by applicant]
Zheyuan Wang and Matthew Gombolay. Learning to Dynamically Coordinate Multi-Robot Teams in Graph Attention Networks. arXiv preprint arXiv:1912.02059, 2019. [cited by applicant]
Keyulu Xu, Weihua Hu, Jure Leskovec, and Stefanie Jegelka. How Powerful are Graph Neural Networks? In International Conference on Learning Representations, 2019. [cited by applicant]
Zhi Yan, Nicolas Jouandeau, and Arab Ali Cherif. A survey and analysis of multi-robot coordination. International Journal of Advanced Robotic Systems, 10(12):399, 2013. [cited by applicant]
Bengio Y, Lodi A, Prouvost A (2020) Machine learning for combinatorial optimization: a methodological tour d'horizon. European Journal of Operational Research. [cited by applicant]
Bogner K, Pferschy U, Unterberger R, Zeiner H (2018) Optimised scheduling in human-robot collaboration-a use case in the assembly of printed circuit boards. International Journal of Production Research 56(16):5522,5540. [cited by applicant]
Casalino A, Zanchettin AM, Piroddi L, Rocco P (2019) Optimal scheduling of human-robot collaborative assembly operations with time petri nets. IEEE Transactions on Automation Science and Engineering Castro E, Petrovic S… [cited by applicant]
Choudhury S, Gupta J, Kochenderfer MJ, Sadigh D, Bohg J (2020) Dynamic multi-robot task allocation under uncertainty and temporal constraints. In: Proceedings of Robotics: Science and Systems (RSS), DOI 10.15607/rss.202… [cited by applicant]
Fout A, Byrd J, Shariat B, Ben-Hur A (2017) Protein interface prediction using graph convolutional networks. In: Advances in neural information processing systems, pp. 6530-6539. [cited by applicant]
Hamaguchi T, Oiwa H, Shimbo M, Matsumoto Y (2017) Knowledge transfer for out-of-knowledge-base entities:A graph neural network approach. arXiv preprint arXiv:170605674. [cited by applicant]
Hamilton WL, Ying R, Leskovec J (2018) Inductive representation learning on large graphs. 1706.02216. [cited by applicant]
Hari SKK, Nayak A, Rathinam S (2020) An approximation algorithm for a task allocation, sequencing and scheduling problem involving a human-robot team. IEEE Robotics and Automation Letters 5(2):2146, 2153. [cited by applicant]
Johnson DB (1977) Efficient algorithms for shortest paths in sparse networks. Journal of the ACM (JACM) 24(1):1,13. [cited by applicant]
Kartal B, Nunes E, Godoy J, Gini M (2016) Monte carlo tree search with branch and bound for multi-robot task allocation. In: The IJCAI-16 workshop on autonomous mobile service robots, vol. 33. [cited by applicant]
Nikou A, Boskos D, Tumova J, Dimarogonas DV (2017) Cooperative planning for coupled multi-agent systems under timed temporal specifications. In: 2017 American Control Conference (ACC), IEEE, pp. 1847,1852. [cited by applicant]
Shiue YR, Lee KC, Su CT (2018) Real-time scheduling for a smart factory using a reinforcement learning approach. Computers & Industrial Engineering 125:604,614. [cited by applicant]
Solovey K, Bandyopadhyay S, Rossi F, Wolf MT, Pavone M (2020) Fast near-optimal heterogeneous task allocation via flow decomposition. arXiv preprintar Xiv:201103603. [cited by applicant]
Tang G, Webb P (2019) Human-robot shared workspace in aerospace factories. Human-robot interaction: safety, standardization, and benchmarking pp. 71-80. [cited by applicant]
Wang H, Chen W, Wang J (2020) Coupled task scheduling for heterogeneous multi-robot system of two robot types performing complex-schedule order fulfillment tasks. Robotics and Autonomous Systems p. 103560. [cited by applicant]
Wang Y, Sun Y, Liu Z, Sarma SE, Bronstein MM, Solomon JM (2019c) Dynamic graph cnn for learning on point clouds. ACM Transactions On Graphics (tog) 38(5):1,12. [cited by applicant]
Wang Z, Gombolay M (2020) Heterogeneous graph attention networks for scalable multi-robot scheduling with temporospatial constraints. In: Robotics: Science and System XVI. [cited by applicant]
Wu Z, Pan S, Chen F, Long G, Zhang C, Philip SY (2020) A comprehensive survey on graph neural networks. IEEE Transactions on Neural Networks and Learning Systems. [cited by applicant]
Yan S, Xiong Y, Lin D (2018) Spatial temporal graph convolutional networks for skeleton-based action recognition. In: Proceedings of the AAAI Conference on Artificial Intelligence, vol. 32, pp. 7444,7452. [cited by applicant]
Yang X, Deng C, Liu T, Tao D (2020) Heterogeneous graph attention network for unsupervised multiple-target domain adaptation. IEEE Transactions on Pattern Analysis and Machine Intelligence. [cited by applicant]
Zhang S, Chen Y, Zhang J, Jia Y (2020) Real-time adaptive assembly scheduling in human-multi-robot collaboration according to human capability. In: 2020 IEEE International Conference on Robotics and Automation (ICRA), I… [cited by applicant]