Seite - 40 - in Joint Austrian Computer Vision and Robotics Workshop 2020

Test Scenario Java Implementation ASPwithinJava Implementation Standalone ASP µ [ms] σ [ms] #TRS µ [ms] σ [ms] #TRS µ [ms] σ [ms] #TRS 1 278.30 226.64 10 121.30 143.50 10 71.90 14.96 10 2 491.30 223.23 10 495.00 339.16 10 278.67 236.67 10 3 2,572.80 4,894.22 10 / / 0 / / 0 Table 3: Runtimeandsolved test runs (TRS) for thedifferent IncubedITassignment implementations Use-case Imperative ASPafter1sec ASPafter5sec ASPafter60sec BMWtaskassign. 7242 3834 3834 3834 BMWpark&chargeassign. 3217 6979 6530 4669 IncubedITassign. 186 293 302 257 Table4: Travelingcosts for all theuse-cases [m] promisingover thequality. To conclude, this work has shown that declarative programming can perform well on real-world logis- ticsscenario,especiallywhenweareinterestedinthe quality of the optimization. Another important ad- vantage of this approach is the separation between the description and the solving of the problem. In fact, performance can be an issue with ASP, espe- ciallyinadynamicplanningscenario,butfortunately state-of-the-artsolverslikeClingoorDLVcomewith manymeta-heuristicsandoptimizations toplaywith. Once the proper settings for the specific scenario are found, solving time can improve considerably, with- out having to modify the code at all. In all the cases in which a greedy algorithm is proven to perform well, in terms of both quality and solving time, im- perative programmingstill remains thebest choice. References [1] Reasoning Web. Semantic Technologies for Intelli- gentDataAccess-9thInternationalSummerSchool 2013, Mannheim, Germany, July 30 - August 2, 2013. Proceedings, volume 8067 of Lecture Notes inComputerScience. Springer, 2013. [2] M. Abseher, M. Gebser, N. Musliu, T. Schaub, and S. Woltran. Shift design with answer set program- ming. Fundam. Inform., 147(1):1–25,2016. [3] C.DodaroandM.Maratea.Nurseschedulingviaan- swersetprogramming. InM.BalducciniandT.Jan- hunen, editors, Logic Programming and Nonmono- tonic Reasoning, pages 301–307. Springer Interna- tionalPublishing, 2017. [4] E. Erdem, E. Aker, and V. Patoglu. Answer set pro- gramming for collaborative housekeeping robotics: representation, reasoning, and execution. 5(4):275– 291,2012. [5] W. Faber, N. Leone, and S. Perri. The intelligent grounder of DLV. In E. Erdem, J. Lee, Y. Lierler, and D. Pearce, editors, Correct Reasoning: Essays on Logic-Based AI in Honour of Vladimir Lifschitz, pages 247–264, Berlin, Heidelberg, 2012. Springer BerlinHeidelberg. [6] F. Fabricius. ASP-based Task Scheduling for Indus- trialTransportRobots. Master’s thesis,GrazUniver- sityofTechnology,2019. [7] M. Gebser, R. Kaminski, B. Kaufmann, and T. Schaub. Answer set solving in practice. Synthe- sis Lectures on Artificial Intelligence and Machine Learning, 6(3):1–238,2012. [8] M. Gebser, R. Kaminski, A. Ko¨nig, and T. Schaub. Advances in gringo series 3. In J. P. Delgrande andW.Faber,editors,LogicProgrammingandNon- monotonic Reasoning, pages 345–351, Berlin, Hei- delberg,2011.SpringerBerlinHeidelberg. [9] M. Gebser, R. Kaminski, and T. Schaub. Ground- ing recursive aggregates: Preliminary report. 2016. Workshopproceeding. [10] M. Gebser, P. Obermeier, T. Schaub, M. Ratsch- Heitmann, and M. Runge. Routing driverless trans- port vehicles in car assembly with answer set pro- gramming. 18(3-4):520–534,2018. [11] M. Gelfond and V. Lifschitz. The stable model se- mantics for logic programming. In R. Kowalski, Bowen, and Kenneth, editors, Proceedings of Inter- national Logic Programming Conference and Sym- posium, pages1070–1080. MITPress, 1988. [12] N. Leone and F. Ricca. Answer set programming: A tour from the basics to advanced development tools and industrial applications. In Reasoning Web. WebLogicRules: 11thInternationalSummerSchool 2015, pages 308–326,072015. [13] V. Nguyen, P. Obermeier, T. C. Son, T. Schaub, and W. Yeoh. Generalized target assignment and path 40