On the relationship between dynamics and complexity in multi-agent collision avoidance

Abstract

This work examines how dynamics and complexity are related in multi-agent collision avoidance. Motivated particularly by work in the field of automated driving, this work considers a variant of the reciprocal n-body collision avoidance problem. In this problem, agents must avoid collision while moving according to individual reward functions in a crowded environment. The main contribution of this work is the result that there is a quantifiable relationship between system dynamics and the requirement for agent coordination, and that this requirement can change the complexity class of the problem dramatically: from P to NEXP or even \(\hbox {NEXP}^{\text {NP}}\). A constructive proof is provided that demonstrates the relationship, and potential practical applications are discussed.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4

Notes

  1. 1.

    Indicator lights are a common channel of communication, but they are notoriously unreliable. Horns also provide a form of communication, but are limited by context. Relative positions and speeds can convey intent, but, as channels of communication, these are very low bandwidth.

  2. 2.

    Section 5 gives an overview of efficient methods for the various types of dynamics computations Problem 1 entails.

  3. 3.

    The problem of choosing when and with whom to communicate, while also difficult, is not a focus of this paper.

  4. 4.

    Breaking this assumption weakens the logical connection in Lemma 3 from a biconditional (if and only if) to a material condition (if).

  5. 5.

    There is even a special tool, the horn, for alerting those around us that someone’s behavior is aberrant.

  6. 6.

    To avoid problems in dealing with dynamic constraints (Wilkie et al. 2009) defined generalized velocity obstacles that are derived in control space rather than velocity space.

References

  1. Abrams, S., & Allen, P. K. (2000). Computing swept volumes. Journal of Visualization and Computer Animation, 11(2), 69–82. https://doi.org/10.1002/1099-1778(200005)11:2%3c69::AID-VIS219%3e3.0.CO;2-7.

    Article  Google Scholar 

  2. Allen, R. E., Clark, A. A., Starek, J. A., & Pavone, M. (2014). A machine learning approach for real-time reachability analysis. In 2014 IEEE/RSJ international conference on intelligent robots and systems (pp. 2202–2208), Chicago, IL, USA, 14–18 September 2014. IEEE. https://doi.org/10.1109/IROS.2014.6942859.

  3. Alonso-Mora, J., Breitenmoser, A., Rufli, M., Beardsley, P. A., & Siegwart, R. (2010). Optimal reciprocal collision avoidance for multiple non-holonomic robots. In A. Martinoli, F. Mondada, N. Correll, G. Mermoud, M. Egerstedt, M. A. Hsieh, L. E. Parker & K. Støy (Eds.), The 10th international symposium distributed autonomous robotic systems, DARS 2010, Lausanne, Switzerland, 1–3 November 2010, Springer tracts in advanced robotics (Vol. 83, pp. 203–216). Springer. https://doi.org/10.1007/978-3-642-32723-0_15.

  4. Alterovitz, R., Siméon, T., & Goldberg, K. Y. (2007). The stochastic motion roadmap: A sampling framework for planning with markov motion uncertainty. In W. Burgard , O. Brock & C. Stachniss (Eds.), Robotics: Science and systems III, 27–30 June 2007, Georgia Institute of Technology, Atlanta, Georgia, USA. The MIT Press. http://www.roboticsproceedings.org/rss03/p30.html.

  5. Bacha, A., Bauman, C., Faruque, R., Fleming, M., Terwelp, C., Reinholtz, C. F., et al. (2008). Odin: Team victortango’s entry in the DARPA urban challenge. Journal of field Robotics, 25(8), 467–492. https://doi.org/10.1002/rob.20248.

    Article  Google Scholar 

  6. Baek, N., Shin, S. Y., & Chwa, K. (1999). On computing translational swept volumes. International Journal of Computational Geometry & Applications, 9(3), 293–317. https://doi.org/10.1142/S0218195999000200.

    MathSciNet  Article  MATH  Google Scholar 

  7. Bekris, K. E. (2010). Avoiding inevitable collision states: Safety and computational efficiency in replanning with sampling-based algorithms. In International conference on robotics and automation (ICRA-10). http://www.cs.rutgers.edu/~kb572/pubs/ics_tradeoffs.pdf. Accessed 15 Feb 2017.

  8. Bekris, K. E., Grady, D. K., Moll, M., & Kavraki, L. E. (2012). Safe distributed motion coordination for second-order systems with different planning cycles. The International Journal of Robotics Research, 31(2), 129–150. https://doi.org/10.1177/0278364911430420.

    Article  Google Scholar 

  9. Bernstein, D. S., Givan, R., Immerman, N., & Zilberstein, S. (2002). The complexity of decentralized control of Markov decision processes. Mathematics of Operations Research, 27(4), 819–840. https://doi.org/10.1287/moor.27.4.819.297.

    MathSciNet  Article  MATH  Google Scholar 

  10. Blass, A., & Gurevich, Y. (1982). On the unique satisfiability problem. Information and Control, 55(1–3), 80–88. https://doi.org/10.1016/S0019-9958(82)90439-9.

    MathSciNet  Article  MATH  Google Scholar 

  11. Boutilier, C. (1996). Planning, learning and coordination in multiagent decision processes. In Y. Shoham (Ed.), Proceedings of the sixth conference on theoretical aspects of rationality and knowledge (pp. 195–210). De Zeeuwse Stromen, The Netherlands, 17–20 March 1996, Morgan Kaufmann.

  12. Chazelle, B., & Dobkin, D. P. (1980). Detection is easier than computation (extended abstract). In R. E. Miller, S. Ginsburg, W. A. Burkhard & R. J. Lipton (Eds.), Proceedings of the 12th annual ACM symposium on theory of computing (pp. 146–153), 28–30 April 1980. Los Angeles, California, USA, ACM. https://doi.org/10.1145/800141.804662.

  13. Daskalakis, K., & Papadimitriou, C. H. (2005). The complexity of games on highly regular graphs. In G. S. Brodal & S. Leonardi (Eds.), 13th annual European symposium algorithms—ESA 2005 (Vol. 3669, pp. 71–82), Palma de Mallorca, Spain, 3–6 October 2005, Proceedings, Springer, Lecture Notes in Computer Science. https://doi.org/10.1007/11561071_9.

  14. Eberly, D. (2008). Intersection of convex objects: The method of separating axes. https://www.geometrictools.com/Documentation/MethodOfSeparatingAxes.pdf. Accessed 2 Jan 2017.

  15. Erickson, L. H., & LaValle, S. M. (2013). A simple, but NP-hard, motion planning problem. In M. desJardins & M. L. Littman (Eds.), Proceedings of the twenty-seventh AAAI conference on artificial intelligence, 14–18 July 2013, Bellevue, Washington, USA. AAAI Press. http://www.aaai.org/ocs/index.php/AAAI/AAAI13/paper/view/6280. Accessed 15 Feb 2017.

  16. Ericson, C. (2005). Real-time collision detection. In Morgan Kaufmann series in interactive 3D technology. Amsterdam: Elsevier. http://opac.inria.fr/record=b1121294

  17. Fiorini, P., & Shiller, Z. (1998). Motion planning in dynamic environments using velocity obstacles. The International Journal of Robotics Research, 17(7), 760–772. https://doi.org/10.1177/027836499801700706.

    Article  Google Scholar 

  18. Fraichard, T., & Asama, H. (2004). Inevitable collision states—A step towards safer robots? Advanced Robotics, 18(10), 1001–1024. https://doi.org/10.1163/1568553042674662.

    Article  Google Scholar 

  19. Gilbert, E. G., Johnson, D. W., & Keerthi, S. S. (1988). A fast procedure for computing the distance between complex objects in three-dimensional space. IEEE Journal on Robotics and Automation, 4(2), 193–203. https://doi.org/10.1109/56.2083.

    Article  Google Scholar 

  20. Goldsmith, J., & Mundhenk, M. (2007). Competition adds complexity. In J. C. Platt, D. Koller, Y. Singer & S. T. Roweis (Eds.), Advances in neural information processing systems 20, Proceedings of the twenty-first annual conference on neural information processing systems (pp. 561–568), Vancouver, British Columbia, Canada, 3–6 December 2007. Curran Associates, Inc. http://papers.nips.cc/paper/3163-competition-adds-complexity.

  21. Halperin, D., & Sharir, M. (1996). A near-quadratic algorithm for planning the motion of a polygon in a polygonal environment. Discrete & Computational Geometry, 16(2), 121–134. https://doi.org/10.1007/BF02716803.

    MathSciNet  Article  MATH  Google Scholar 

  22. Harding, J., Powell, G., Yoon, R., Fikentscher, J., Doyle, C., Sade, D., Lukuc, M., Simons, J., & Wang, J. (2014). Vehicle-to-vehicle communications: Readiness of V2V technology for application. Technical Report DOT HS 812 014, U.S. Department of Transportation, National Highway Transportation Safety Administration, National Highway Traffic Safety Administration, 1200 New Jersey Avenue SE. Washington, DC 20590. http://www.nhtsa.gov/staticfiles/rulemaking/pdf/V2V/Readiness-of-V2V-Technology-for-Application-812014.pdf. Accessed 15 Feb 2017.

  23. Hauser, K. K. (2012). The minimum constraint removal problem with three robotics applications. In E. Frazzoli, T. Lozano-Pérez, N. Roy & D. Rus (Eds.), Algorithmic foundations of robotics X-proceedings of the tenth workshop on the algorithmic foundations of robotics, WAFR 2012 (Vol. 86, pp. 1–17). MIT, Cambridge, Massachusetts, USA, 13–15 June 2012. Springer Tracts in Advanced Robotics. Springer. https://doi.org/10.1007/978-3-642-36279-8_1.

  24. Hopcroft, J., Schwartz, J., & Sharir, M. (1984). On the complexity of motion planning for multiple independent objects; PSPACE-hardness of the “Warehouseman’s Problem”. The International Journal of Robotics Research, 3(4), 76–88. http://ijr.sagepub.com/content/3/4/76.short.

  25. Hornung, A., Wurm, K. M., Bennewitz, M., Stachniss, C., & Burgard, W. (2013). OctoMap: An efficient probabilistic 3D mapping framework based on octrees. Autonomous Robots. https://doi.org/10.1007/s10514-012-9321-0, http://octomap.github.com, software available at http://octomap.github.com.

  26. Jansson, J. (2005). Collision avoidance theory: with application to automotive collision mitigation. Linköping studies in science and technology: Dissertations, Department of Electrical Enginering, University. https://books.google.com/books?id=ik8wNQAACAAJ.

  27. Jiménez, P., Thomas, F., & Torras, C. (2001). 3D collision detection: A survey. Computers & Graphics, 25(2), 269–285. https://doi.org/10.1016/S0097-8493(00)00130-8.

    Article  Google Scholar 

  28. Johnson, J. K. (2016). A novel relationship between dynamics and complexity in multi-agent collision avoidance. In Proceedings of robotics: Science and systems. Ann Arbor, Michigan. https://doi.org/10.15607/RSS.2016.XII.030.

  29. Johnson, J., & Hauser, K. K. (2012). Optimal acceleration-bounded trajectory planning in dynamic environments along a specified path. In IEEE international conference on robotics and automation, ICRA 2012 (pp. 2035–2041), 14–18 May 2012, St. Paul, Minnesota, USA. IEEE. https://doi.org/10.1109/ICRA.2012.6225233.

  30. Kaelbling, L. P., & Lozano-Pérez, T. (2013). Integrated task and motion planning in belief space. The International Journal of Robotics Research, 32(9–10), 1194–1227. https://doi.org/10.1177/0278364913484072.

    Article  Google Scholar 

  31. Kamat, V. V. (1993). A survey of techniques for simulation of dynamic collision detection and response. Computers & Graphics, 17(4), 379–385. https://doi.org/10.1016/0097-8493(93)90024-4.

    Article  Google Scholar 

  32. Kambhampati, S., Cutkosky, M. R., Tenenbaum, M., & Lee, S. H. (1991). Combining specialized reasoners and general purpose planners: A case study. In T. L. Dean & K. McKeown (Eds.), Proceedings of the 9th national conference on artificial intelligence, Anaheim, CA, USA (Vol. 1, pp. 199–205), 14–19 July 1991. AAAI Press/The MIT Press. http://www.aaai.org/Library/AAAI/1991/aaai91-032.php.

  33. Kim, Y. J., Varadhan, G., Lin, M. C., & Manocha, D. (2004). Fast swept volume approximation of complex polyhedral models. Computer-Aided Design, 36(11), 1013–1027. https://doi.org/10.1016/j.cad.2004.01.004.

    Article  Google Scholar 

  34. LaValle, S. M. (2006). Planning algorithms. Cambridge: Cambridge University Press.

    Google Scholar 

  35. LaValle, S. M., & Kuffner, J. J. (2001). Randomized kinodynamic planning. The International Journal of Robotics Research, 20(5), 378–400. https://doi.org/10.1177/02783640122067453.

    Article  Google Scholar 

  36. Lin, M. C., & Canny, J. F. (1991). A Fast Algorithm for incremental distance calculation. In IEEE international conference on robotics and automation (pp. 1008–1014). http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.30.2174. Accessed 15 Feb 2017.

  37. Littman, M. L., Dean, T. L., & Kaelbling, L. P. (1995) On the complexity of solving Markov decision problems. In UAI ’95: Proceedings of the eleventh annual conference on uncertainty in artificial intelligence (pp. 394–402), Montreal, Quebec, Canada, 18–20 August 1995. https://dslpitt.org/uai/displayArticleDetails.jsp?mmnu=1&smnu=2&article_id=457&proceeding_id=11. Accessed 15 Feb 2017.

  38. Mazer, E., Ahuactzin, J. M., & Bessière, P. (1998). The ariadne’s clew algorithm. Journal of Artificial Intelligence Research (JAIR), 9, 295–316. https://doi.org/10.1613/jair.468.

    Article  MATH  Google Scholar 

  39. Mitter, S., & Sahai, A. (1999). Information and control: Witsenhausen revisited (pp. 281–293). London: Springer. https://doi.org/10.1007/BFb0109735.

    Google Scholar 

  40. Montemerlo, M., Becker, J., Bhat, S., Dahlkamp, H., Dolgov, D., Ettinger, S., et al. (2008). Junior: The stanford entry in the urban challenge. Journal of field Robotics, 25(9), 569–597. https://doi.org/10.1002/rob.20258.

    Article  Google Scholar 

  41. Ó’Dúnlaing, C. (1987). Motion planning with inertial constraints. Algorithmica, 2, 431–475. https://doi.org/10.1007/BF01840370.

    MathSciNet  Article  MATH  Google Scholar 

  42. Oliehoek, F. A., Witwicki, S. J., & Kaelbling, L. P. (2012). Influence-based abstraction for multiagent systems. In Proceedings of the twenty-sixth AAAI conference on artificial intelligence, 22–26 July 2012, Toronto, Ontario, Canada. http://www.aaai.org/ocs/index.php/AAAI/AAAI12/paper/view/5047.

  43. Paden, B., Cáp, M., Yong, S. Z., Yershov, D. S., & Frazzoli, E. (2016). A survey of motion planning and control techniques for self-driving urban vehicles. CoRR http://arxiv.org/abs/1604.07446.

  44. Papadimitriou, C. H., & Tsitsiklis, J. N. (1987). The complexity of markov decision processes. Mathematics of Operations Research, 12(3), 441–450. http://www.jstor.org/stable/3689975

  45. Petti, S., & Fraichard, T. (2005). Safe motion planning in dynamic environments. In 2005 IEEE/RSJ international conference on intelligent robots and systems (pp. 2210–2215), Edmonton, Alberta, Canada, 2–6 August 2005. IEEE. https://doi.org/10.1109/IROS.2005.1545549.

  46. Reif, J. H. (1979). Complexity of the mover’s problem and generalizations (extended abstract). In 20th annual symposium on foundations of computer science (pp. 421–427), San Juan, Puerto Rico, 29–31 October 1979. IEEE Computer Society. https://doi.org/10.1109/SFCS.1979.10.

  47. Reif, J. H., & Sharir, M.(1985). Motion planning in the presence of moving obstacles. In 26th annual symposium on foundations of computer science (pp. 144–154), Portland, Oregon, USA, 21–23 October 1985. IEEE Computer Society. https://doi.org/10.1109/SFCS.1985.36.

  48. Schoettle, B., & Sivak, M. (2015). A preliminary analysis of real-world crashes involving self-driving vehicles. Technical report, The University of Michigan Transportation Research Institute.

  49. Shamos, M. (1978). Computational geometry. Ph.D. thesis, Yale University.

  50. Shiller, Z., Gal, O., & Fraichard, T. (2010). The nonlinear velocity obstacle revisited: The optimal time horizon. In Guaranteeing safe navigation in dynamic environments workshop, Anchorage, United States. https://hal.inria.fr/inria-00562249. Accessed 15 Feb 2017.

  51. Shoham, Y., & Tennenholtz, M. (1995). On social laws for artificial agent societies: Off-line design. Artificial Intelligence, 73(1–2), 231–252. https://doi.org/10.1016/0004-3702(94)00007-N.

    Article  Google Scholar 

  52. Solovey, K., & Halperin, D. (2015). On the hardness of unlabeled multi-robot motion planning. In Robotics: Science and systems XI, Sapienza University of Rome, Rome, Italy, 13–17 July 2015. http://www.roboticsproceedings.org/rss11/p46.html.

  53. Täubig, H., Bäuml, B., & Frese, U. (2011). Real-time swept volume and distance computation for self collision detection. In 2011 IEEE/RSJ international conference on intelligent robots and systems, IROS 2011 (pp. 1585–1592), San Francisco, CA, USA, 25–30 September 2011. IEEE. https://doi.org/10.1109/IROS.2011.6094611.

  54. Urmson, C., Anhalt, J., Bagnell, J. A. D., Baker, C. R., Bittner, R. E., Dolan, J. M., Duggins, D., Ferguson, D., Galatali, T., Geyer, H., Gittleman, M., Harbaugh, S., Hebert, M., Howard, T., Kelly, A., Kohanbash, D., Likhachev, M., Miller, N., Peterson, K., Rajkumar, R., Rybski, P., Salesky, B., Scherer, S., Seo, Y. W., Simmons, R,, Singh, S., Snider, J. M., Stentz, A. T., Whittaker, W. R. L., & Ziglar, J. (2007). Tartan racing: A multi-modal approach to the DARPA urban challenge. Technical Report CMU-RI-TR-, Robotics Institute, Pittsburgh, PA. http://www.ri.cmu.edu/publication_view.html?pub_id=6906. Accessed 15 Feb 2017.

  55. van den Berg, J., Guy, S. J., Lin, M. C., & Manocha, D. (2009). Reciprocal n-body collision avoidance. In C. Pradalier, R. Siegwart & G. Hirzinger (Eds.), The 14th international symposium robotics research, ISRR 2009, August 31–September 3, 2009, Lucerne, Switzerland, Springer tracts in advanced robotics (Vol. 70, pp. 3–19). Springer. https://doi.org/10.1007/978-3-642-19457-3_1.

  56. van den Berg J. P., Snape, J., Guy, S. J., & Manocha, D. (2011). Reciprocal collision avoidance with acceleration-velocity obstacles. In IEEE international conference on robotics and automation, ICRA 2011, Shanghai, China, 9–13 May 2011 (pp. 3475–3482). IEEE. https://doi.org/10.1109/ICRA.2011.5980408.

  57. Valtazanos, A., & Ramamoorthy, S. (2011). Online motion planning for multi-robot interaction using composable reachable sets. In T. Röfer, N. M. Mayer, J. Savage, U. Saranli (Eds.), RoboCup 2011: Robot Soccer World Cup XV [papers from the 15th annual RoboCup international symposium, Istanbul, Turkey, July 2011]. Lecture Notes in Computer Science (Vol. 7416, pp. 186–197). Springer. https://doi.org/10.1007/978-3-642-32060-6_16.

  58. von Dziegielewski, A., Erbes, R., & Schömer, E. (2010). Conservative swept volume boundary approximation. In G. Elber, A. Fischer, J. Keyser & M. Kim (Eds.), ACM symposium on solid and physical modeling, Proceedings of the 14th ACM symposium on solid and physical modeling, SPM 2010 (pp. 171–176), Haifa, Israel, 1–3 September 2010. ACM. https://doi.org/10.1145/1839778.1839804.

  59. Weller, R. (2013). New geometric data structures for collision detection and haptics. Springer series on touch and haptic systems. Springer International Publishing. http://www.springer.com/computer/theoretical+computer+science/book/978-3-319-01019-9. Accessed 15 Feb 2017.

  60. Wilkerson, J. L., Bobinchak, J., Culp, M., Clark, J., Halpin-Chan, T., Estabridis, K., & Hewer, G. (2014). Two-dimensional distributed velocity collision avoidance. Technical Report NAWCWD TP 8786, Physics Division, Research and Intelligence Department, Naval Air Warfare Center Weapons Division, China Lake, CA 93555-6100. http://www.dtic.mil/dtic/tr/fulltext/u2/a598520.pdf. Accessed 15 Feb 2017.

  61. Wilkie, D., van den Berg, J. P., & Manocha, D. (2009). Generalized velocity obstacles. In 2009 IEEE/RSJ international conference on intelligent robots and systems (pp. 5573–5578), 11–15 October 2009, St. Louis, MO, USA. IEEE. https://doi.org/10.1109/IROS.2009.5354175.

  62. Witsenhausen, H. S. (1968). A counterexample in stochastic optimum control. SIAM Journal on Control, 6(1), 131–147. https://doi.org/10.1137/0306011.

    MathSciNet  Article  MATH  Google Scholar 

  63. Ziegler, J., Bender, P., Dang, T., & Stiller, C. (2014). Trajectory planning for bertha—A local, continuous method. In 2014 IEEE intelligent vehicles symposium proceedings (pp. 450–457), Dearborn, MI, USA, 8–11 June 2014. IEEE. https://doi.org/10.1109/IVS.2014.6856581.

Download references

Acknowledgements

The author would like to acknowledge the reviewers for providing many insightful comments and corrections, as well as Philipp Robbel and Elmar Mair for comments and feedback, and Valerie Aquila for assistance with figures and editing.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Jeffrey Kane Johnson.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

This is one of several papers published in Autonomous Robots comprising the “Special Issue on Robotics Science and Systems”.

Appendix

Appendix

The first portion of the appendix will prove that the VO representation cannot conservatively approximate inertially constrained systems. It will also show that the VO representation belongs to the family of ICS representations using the inevitable collision obstacle (ICO) concept.

The last portion of the appendix will describe a conjecture about the problem complexity of finding a unique set of, collision-free, non-disjoint SPs.

A.1 The inevitable collision obstacle

Definition 15

An inevitable collision obstacle (ICO) is the set of states of an agent \(A\) that result in collision with \(\mathcal {B}_i\) for any control sequence \(\phi \) is applied to \(A\):

$$\begin{aligned} \text {ICO}(\mathcal {B}_i)=\{\mathbf{x}\;|\;\forall \phi ,\exists t::A(\phi (\mathbf{x},t))\cap \mathcal {B}_i\ne \emptyset \} \end{aligned}$$

The ICO is closely related to the ICS concept, both of which were introduced by Fraichard and Asama (2004).

A.2 The velocity obstacle

This section recalls the velocity obstacle and relevant properties. We use the definitions from Fiorini and Shiller (1998), and the reader is referred to that work for more detail.Footnote 6

In this section, assume \(t\in T\), where \(T=[0,\infty )\) is a finite time horizon. Let \(\varPhi _v\) be the set of feasible velocity commands for \(A\), and let \(\phi _v(\mathbf{x}, t)\) denote the state of \(A\) after constant velocity v is applied to initial state \(\mathbf{x}\) for a time t.

Definition 16

The velocity obstacle for \(A\) due to \(O_i\), written \(\text {VO}_{A|O_i}\), is the set of velocities such that \(A\) at some point enters into a collision state with \(O\). In other words, given initial state \(\mathbf{x}\), and for all feasible velocity commands \(v\in \varPhi _v\) there is a collision at some time \(t\in T\) between \(A(\mathbf{x})\) and the state space obstacle \(\mathcal {B}_i\) due to \(O_i\):

$$\begin{aligned} \text {VO}_{A|O_i}=\{v\;|\;\exists t::A(\phi _v(\mathbf{x},t))\cap \mathcal {B}_i\ne \emptyset \} \end{aligned}$$

A.3 Velocity obstacles and inertially constrained systems

Lemma 9

The VO representation cannot guarantee collision avoidance in inertially constrained systems.

Proof

By Definition 16, the complement of the velocity obstacle is exactly the set of all velocities that, when instantaneously applied, would avoid collision. However, controlling to a velocity instantaneously is impossible in an inertially constrained system. Therefore, the complement of the velocity obstacle is unreachable, and by Lemma 1, it cannot be used to guarantee non-collision. \(\square \)

Velocity obstacle and inevitable collision obstacle equivalence

The reader will note the similarity between Definition 15 and 16, and work by Shiller et al. (2010) suggests that a deeper relationship exists. The proof proceeds by exploiting the similarity and showing that ICO computations are both necessary and sufficient in order to compute the VO.

Definition 17

A velocity \(\text {ICO}\) (ICO\(_v\)) for a given state space obstacle \(\mathcal {B}_i\) is an \(\text {ICO}\) computed over the velocity control trajectory set \(\varPhi _v\):

$$\begin{aligned} \text {ICO}(\mathcal {B}_i)_v=\{\mathbf{x}\;|\forall \phi _v,\exists t::A(\phi _v(\mathbf{x},t))\cap \mathcal {B}_i\ne \emptyset \} \end{aligned}$$

Lemma 10

Computing a velocity obstacle is exactly equivalent to computing an inevitable collision obstacle over a restricted control space.

Proof

For a given obstacle \(O_i\) and corresponding state space obstacle \(\mathcal {B}_i\), use Definition 16 to perform a variable rewrite on the definition of a velocity ICO (Definition 17):

$$\begin{aligned} \text {ICO}(\mathcal {B}_i)_v&=\{\mathbf{x}\;|\;\forall \phi _v,\exists t::A(\phi _v(\mathbf{x},t))\cap \mathcal {B}_i\ne \emptyset \}\\&=\{\mathbf{x}\;|\;\forall \phi _v, v\in \text {VO}_{A|O_i}\} \end{aligned}$$

Thus, the ICO(\(\mathcal {B})_v\) and \(\hbox {VO}_{A|O}\) are equivalent, which means that the velocity obstacle representation is equivalent to the ICO representation over a restricted control space. \(\square \)

The result of Lemma 10 provides a simple but formal unification of two common techniques for collision avoidance under the same theoretical framework: that velocity obstacles are exactly inevitable collision obstacles over a restricted set of the inputs.

A.4 A special case of coordination

The proof of Lemma 3 asserts through Definition 3 that finding a unique set of collision-free, non-disjoint SPs induces a coordination requirement. Invoking this type of coordination has interesting complexity implications because the general problem of identifying a unique assignment of such SPs is likely reducible to a Unique-SAT problem, which is coNP-Hard (Blass and Gurevich 1982). The following conjecture captures this:

Conjecture 2

There is no efficiently computable (i.e. P-time) solution to identifying a unique set of collision-free stopping paths in a system that does not exhibit SP disjointness.

Investigation of Conjecture 2 would be an interesting point for future work.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Johnson, J.K. On the relationship between dynamics and complexity in multi-agent collision avoidance. Auton Robot 42, 1389–1404 (2018). https://doi.org/10.1007/s10514-018-9743-4

Download citation

Keywords

  • Complexity
  • Dynamics
  • Collision avoidance