Abstract

Autonomous manipulation is a challenging problem in field robotics due to uncertainty in object properties, constraints, and coupling phenomenon with robot control systems. Humans learn motion primitives over time to effectively interact with the environment. We postulate that autonomous manipulation can be enabled by basic sets of motion primitives as well, but do not necessitate mimicking human motion primitives. This work presents an approach to generalized optimal motion primitives using physics-informed neural networks. Our simulated and experimental results demonstrate that optimality is notionally maintained where the mean maximum observed final position percent error was 0.564% and the average mean error for all the trajectories was 1.53%. These results indicate that notional generalization is attained using a physics-informed neural network approach that enables near optimal real-time adaptation of primitive motion profiles.

Issue Section:
Research Papers
Keywords:
machine learning, robotics, optimization algorithms, neural networks
Topics:
Artificial neural networks, Physics, Robotics, Robots, Machine learning, Trajectories (Physics), Errors

References

1.
Hogan
,
N.
,
2017
, “Physical Interaction via Dynamic Primitives,”
Geometric and Numerical Foundations of Movements
,
J. P.
Laumond
,
N.
Mansard
, and
J. B.
Lasserre
, eds., Springer Tracts in Advanced Robotics, vol. 117, Springer, Cham, pp.
269
–
299
.
Google Scholar
Crossref
Search ADS
 
2.
Slightam
,
J. E.
,
McArthur
,
D. R.
,
Spencer
,
S. J.
, and
Buerger
,
S. P.
,
2021
, “Passivity Analysis of Quadrotor Aircraft for Physical Interactions,”
2021 Aerial Robotic Systems Physically Interacting with the Environment (AIRPHARO)
,
Biograd na Moru, Croatia
,
IEEE
, pp.
1
–
8
.
Crossref
Search ADS
 
3.
Slightam
,
J. E.
,
Nagurka
,
M. L.
, and
Barth
,
E. J.
,
2018
, “
Sliding Mode Impedance Control of a Hydraulic Artificial Muscle
,”
2018 Dynamic Systems and Control Conference
,
Atlanta, GA
.
Sept. 30–Oct. 3
,
ASME
, p. V001T13A003.
Google Scholar
Crossref
Search ADS
 
4.
Slightam
,
J. E.
,
Barth
,
E. J.
, and
Nagurka
,
M. L.
,
2019
, “
Sliding Mode Impedance and Stiffness Control of a Pneumatic Cylinder
,”
2019 Dynamic Systems and Control Conference
,
Park City, UT
,
Oct. 8–11
,
ASME
, p. V001T02A002.
Google Scholar
Crossref
Search ADS
 
5.
Settimi
,
A.
,
Caporale
,
D.
,
Kryczka
,
P.
,
Ferrati
,
M.
, and
Pallottino
,
L.
,
2016
, “
Motion Primitive Based Random Planning for Loco-Manipulation Tasks
,”
2016 IEEE-RAS 16th International Conference on Humanoid Robots (Humanoids)
,
Cancun, Mexico
,
Nov. 15–17
,
IEEE
, pp.
1059
–
1066
.
Google Scholar
Crossref
Search ADS
 
6.
Cohen
,
B. J.
,
Chitta
,
S.
, and
Likhachev
,
M.
,
2010
, “
Search-Based Planning for Manipulation With Motion Primitives
,”
2010 IEEE International Conference on Robotics and Automation
,
Anchorage, AK
,
May 3–7
,
IEEE
, pp.
2902
–
2908
.
Google Scholar
Crossref
Search ADS
 
7.
Stulp
,
F.
,
Theodorou
,
E. A.
,
Kalakrishnan
,
M.
,
Pastor
,
P.
,
Righetti
,
L.
, and
Schaal
,
S.
,
2011
, “
Learning Motion Primitive Goals for Robust Manipulation
,”
2011 IEEE/RSJ International Conference on Intelligent Robots and Systems
,
San Francisco, CA
,
Sept. 25–30
, pp.
325
–
331
. http:/dx.doi.org/10.1109/IROS.2011.6094877
Google Scholar
Crossref
Search ADS
 
8.
Stulp
,
F.
,
Theodorou
,
E.
,
Kalakrishnan
,
M.
,
Pastor
,
P.
,
Righetti
,
L.
, and
Schaal
,
S.
,
2011
, “
Learning Motion Primitive Goals for Robust Manipulation
,” 2011 IEEE, RSJ International Conference on Intelligent Robots and Systems,
IEEE
, pp.
325
–
331
.
9.
Saxena
,
D. M.
,
Saleem
,
M. S.
, and
Likhachev
,
M.
,
2021
, “
Manipulation Planning Among Movable Obstacles Using Physics-Based Adaptive Motion Primitives
,”
2021 IEEE International Conference on Robotics and Automation (ICRA)
,
Xi'an, China
,
May 30–June 5
,
IEEE
, pp.
6570
–
6576
.
Google Scholar
Crossref
Search ADS
 
10.
Cohen
,
B. J.
,
Subramania
,
G.
,
Chitta
,
S.
, and
Likhachev
,
M.
,
2011
, “
Planning for Manipulation With Adaptive Motion Primitives
,”
2011 IEEE International Conference on Robotics and Automation
,
Shanghai, China
,
May 9–13
,
IEEE
, pp.
5478
–
5485
.
Google Scholar
Crossref
Search ADS
 
11.
Lutter
,
M.
, and
Peters
,
J.
,
2023
, “
Combining Physics and Deep Learning to Learn Continuous-time Dynamics Models
,”
Int. J. Rob. Res.
,
42
(
3
), pp.
83
–
107
.
Google Scholar
Crossref
Search ADS
 
12.
Lutter
,
M.
,
Listmann
,
K.
, and
Peters
,
J.
,
2019
, “
Deep Lagrangian Networks for End-to-End Learning of Energy-Based Control for Under-Actuated Systems
,”
2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
,
Macau, China
,
Nov. 3–8
,
IEEE
, pp.
7718
–
7725
.
Google Scholar
Crossref
Search ADS
 
13.
Lutter
,
M.
,
Ritter
,
C.
, and
Peters
,
J.
,
2019
, “
Deep Lagrangian Networks: Using Physics as Model Prior for Deep Learning
,” arXiv preprint arXiv:1907.04490.
14.
Duong
,
T.
,
Altawaitan
,
A.
,
Stanley
,
J.
, and
Atanasov
,
N.
,
2024
, “
Port-Hamiltonian Neural ODE Networks on Lie Groups for Robot Dynamics Learning and Control
,”
IEEE Trans. Rob.
,
40
, pp.
3695
–
3715
.
Google Scholar
Crossref
Search ADS
 
15.
Lee
,
K.
,
Trask
,
N.
, and
Stinis
,
P.
,
2021
, “
Machine Learning Structure Preserving Brackets for Forecasting Irreversible Processes
,”
Adv. Neural Inf. Process. Syst.
,
34
, pp.
5696
–
5707
.
16.
Zhong
,
Y. D.
,
Dey
,
B.
, and
Chakraborty
,
A.
,
2020
, “
Dissipative SymODEN: Encoding Hamiltonian Dynamics With Dissipation and Control Into Deep Learning
,
International Conference on Learning Representations
,
Virtual
,
Apr 26– May 1
.
17.
Shima
,
R.
,
Okura
,
Y.
,
Fujimoto
,
K.
, and
Maruta
,
I.
,
2019
, “
On Path Following Control of Nonholonomic Port-Hamiltonian Systems via Generalized Canonical Transformations
,”
IFAC-PapersOnLine
,
52
(
16
), pp.
298
–
303
.
Google Scholar
Crossref
Search ADS
 
18.
Jin
,
W.
,
Wang
,
Z.
,
Yang
,
Z.
, and
Mou
,
S.
,
2020
, “
Pontryagin Differentiable Programming: An End-to-End Learning and Control Framework
,”
Adv. Neural Inf. Process. Syst.
,
33
, pp.
7979
–
7992
.
19.
Slightam
,
J. E.
, and
Griego
,
A. D.
,
2023
, “
Deep Neural Network Design for Improving Stability and Transient Behavior in Impedance Control Applications
,”
2023 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM)
,
Seattle, WA
,
June 28–30
,
IEEE/ASME
, pp.
1336
–
1343
.
Google Scholar
Crossref
Search ADS
 
20.
Young
,
C.
,
Stahoviak
,
C.
,
Kim
,
R.
, and
Slightam
,
J. E.
,
2024
, “
Rapid Constrained Object Motion Estimation Based on Centroid Localization of Semantically Labeled Objects
,”
2024 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM)
,
Boston, MA
,
IEEE/ASME
, pp.
791
–
798
.
Google Scholar
Crossref
Search ADS
 
21.
Bryson
,
A. E., Jr.
, and
Ho
,
Y. -C.
,
1975
,
Applied Optimal Control: Optimization, Estimation, and Control
,
Taylor & Francis Group
,
New York
.
22.
Beaver
,
L. E.
, and
Malikopoulos
,
A. A.
,
2024
, “
Optimal Control of Differentially Flat Systems is Surprisingly Easy
,”
Automatica
,
159
, p.
111404
.
Google Scholar
Crossref
Search ADS
 
23.
Ross
,
I. M.
,
2015
,
A Primer on Pontryagin’s Principle in Optimal Control
, 2nd ed.,
Collegiate Publishers
,
San Francisco, CA
.
Copyright © 2024 by ASME
You do not currently have access to this content.