Personal and Ubiquitous Computing

, Volume 18, Issue 1, pp 205–221 | Cite as

Activity recognition for creatures of habit

Energy-efficient embedded classification using prediction
  • Dawud GordonEmail author
  • Jürgen Czerny
  • Michael Beigl
Original Article


Energy storage is quickly becoming the limiting factor in mobile pervasive technology. We introduce a novel method for activity recognition which leverages the predictability of human behavior to conserve energy by dynamically selecting sensors. We further present a taxonomy of existing approaches to dynamically reducing consumption while maintaining recognition rates. The novel algorithm conserves energy by quantifying activity-sensor dependencies and using prediction methods to identify likely future activities. The approach is implemented and simulated using two activity recognition data sets, and the effects of the novel method are evaluated in terms of recognition rates, energy consumption, and prediction rates. The results indicate that switching off sensors only significantly affects prediction under extreme conditions and that these effects can be counteracted by adjusting system parameters. Large savings in energy can be achieved at very low cost, for example, recognition losses of 1.5 pp with 84.8 % energy savings for the first data set, and 2.8 pp and 89.9 % for the second.


Feature Vector Hide Markov Model Energy Saving Recognition Rate Activity Recognition 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 1.
    Au L, Bui A, Batalin M, Kaiser W (2012) Energy-efficient context classification with dynamic sensor control. Biomed Circ Syst IEEE Trans 6(2):167–178CrossRefGoogle Scholar
  2. 2.
    Benbasat AY, Paradiso JA (2007) A framework for the automated generation of power-efficient classifiers for embedded sensor nodes. In: SenSys ’07: proceedings of the 5th international conference on embedded networked sensor systems. ACM, New York, pp 219–232Google Scholar
  3. 3.
    Berchtold M, Budde M, Gordon D, Schmidtke H, Beigl M (2010) Actiserv: activity recognition service for mobile phones. In: Wearable Computers (ISWC), 2010 international symposium, pp 1–8Google Scholar
  4. 4.
    Bharatula N, Stager M, Lukowicz P, Tröster G (2005) Power and size optimized multi-sensor context recognition platform. In: Wearable Computers, 2005. Proceedings. Ninth IEEE international symposium, pp 194–195Google Scholar
  5. 5.
    Bharatula NB, Ossevoort S, Stäger M, Tröster G (2004) Towards wearable autonomous microsystems. In: Pervasive, pp 225–237Google Scholar
  6. 6.
    Bharatula NB, Stäger M, Lukowicz P, Tröster G (2005) Empirical study of design choices in multi-sensor context recognition systems. In: IFAWC: 2nd international forum on applied wearable computing, pp 79–93Google Scholar
  7. 7.
    Bishop CM (2006) Pattern recognition and machine learning, 1st ed. 2006. corr. 2nd printing edn. Springer, BerlinGoogle Scholar
  8. 8.
    Bouten C, Koekkoek K, Verduin M, Kodde R, Janssen J (1997) A triaxial accelerometer and portable data processing unit for the assessment of daily physical activity. Biomed Eng IEEE Trans 44(3):136–147CrossRefGoogle Scholar
  9. 9.
    Cakmakci O, Coutaz J, Laerhoven KV, Werner Gellersen H (2002) Context awareness in systems with limited resources. In: In Proceeding of the third workshop on artificial intelligence in mobile systems (AIMS), ECAI 2002, pp 21–29Google Scholar
  10. 10.
    Gao L, Bourke A, Nelson J (2012) Activity recognition using dynamic multiple sensor fusion in body sensor networks. In: Engineering in Medicine and Biology Society (EMBC), 2012 annual international conference of the IEEE, pp 1077–1080Google Scholar
  11. 11.
    Gordon D, Czerny J, Miyaki T, Beigl M (2012) Energy-efficient activity recognition using prediction. In: Wearable Computers (ISWC), 2012 16th international symposium, pp 29–36Google Scholar
  12. 12.
    Gordon D, Hanne JH, Berchtold M, Miyaki T, Beigl M (2012) Recognizing group activities using wearable sensors. In: Puiatti A, Gu T (eds) Mobile and ubiquitous systems: computing, networking, and services, lecture notes of the institute for computer sciences, social informatics and telecommunications engineering. Springer, Berlin, vol 104, pp 350–361Google Scholar
  13. 13.
    Gordon D, Hanne JH, Berchtold M, Shirehjini A, Beigl M (2012) Towards collaborative group activity recognition using mobile devices. Mobile Networks and Applications pp 1–15Google Scholar
  14. 14.
    Gordon D, Schmidtke H, Beigl M, von Zengen G (2010) A novel micro-vibration sensor for activity recognition: potential and limitations. In: Wearable Computers (ISWC), 2010 international symposium, pp 1–8Google Scholar
  15. 15.
    Gordon D, Sigg S, Ding Y, Beigl M (2011) Using prediction to conserve energy in recognition on mobile devices. In: Pervasive computing and communications workshops (PERCOM Workshops), 2011 IEEE international conference, pp 364–367Google Scholar
  16. 16.
    Könönen V, Mäntyjärvi J, Similä H, Pärkkä J, Ermes M (2010) Automatic feature selection for context recognition in mobile devices. Pervasive Mob Comput 6(2):181–197CrossRefGoogle Scholar
  17. 17.
    Krause A, Ihmig M, Rankin E, Leong D, Gupta S, Siewiorek D, Smailagic A, Deisher M, Sengupta U (2005) Trading off prediction accuracy and power consumption for context-aware wearable computing. In: Wearable Computers, 2005. Proceedings. Ninth IEEE international symposium, pp 20–26Google Scholar
  18. 18.
    Lin K (2010) Energy-accuracy aware localization for mobile devices. In: Proceedings of the 8th international conference on mobile systems applications and services MobiSys 10. In Mobisys, ACM Press, p 285Google Scholar
  19. 19.
    Lu H, Brush AJB, Priyantha B, Karlson AK, Liu J (2011) SpeakerSense: energy efficient unobtrusive speaker identification on mobile phones. Pervasive Comput 6696:188–205Google Scholar
  20. 20.
    Mayrhofer R, Radi H, Ferscha A (2003) Recognizing and predicting context by learning from user behavior. In: Kotsis WSG, Ferscha A, Ibrahim K (eds) Proceedings MoMM 2003: 1st international conference on advances in mobile multimedia. Austrian Computer Society (OCG), vol 171, pp 25–35Google Scholar
  21. 21.
    Paek J, Kim J, Govindan R (2010) Energy-efficient rate-adaptive GPS-based positioning for smartphones. In: Proceedings of the 8th international conference on Mobile systems applications and services MobiSys 10, MobiSys 10. ACM, ACM Press, p 299Google Scholar
  22. 22.
    Rabiner L (1989) A tutorial on hidden markov models and selected applications in speech recognition. Proceedings of the IEEE 77(2):257–286CrossRefGoogle Scholar
  23. 23.
    Raffa G, Lee J, Nachman L, Song J (2010) Don’t slow me down: bringing energy efficiency to continuous gesture recognition. In: Wearable Computers ISWC 2010 international symposium. IEEE, pp 1–8Google Scholar
  24. 24.
    Roggen D, Calatroni A, Rossi M, Holleczek T, Forster K, Tröster G, Lukowicz P, Bannach D, Pirkl G, Ferscha A, Doppler J, Holzmann C, Kurz M, Holl G, Chavarriaga R, Sagha H, Bayati H, Creatura M, del R Millan J (2010) Collecting complex activity datasets in highly rich networked sensor environments. In: Networked sensing systems (INSS), 2010 Seventh international conference, pp 233–240Google Scholar
  25. 25.
    Roy N, Misra A, Julien C, Das SK, Biswas J (2011) An energy-efficient quality adaptive framework for multi-modal sensor context recognition. In: IEEE international conference on pervasive computing and communications PerCom. Institute for Infocomm Research, Singapore, IEEE, pp 63–73Google Scholar
  26. 26.
    Sigg S, Gordon D, Zengen G, Beigl M, Haseloff S, David K (2011) Investigation of context prediction accuracy for different context abstraction levels. Mobile Computing, IEEE Trans (99):1Google Scholar
  27. 27.
    Stäger M, Lukowicz P, Tröster G (2004) Implementation and evaluation of a low-power sound-based user activity recognition system. In: ISWC ’04: Proceedings of the eighth international symposium on wearable computers. IEEE Computer Society, Washington, pp 138–141Google Scholar
  28. 28.
    Stäger M, Lukowicz P, Tröster G (2007) Power and accuracy trade-offs in sound-based context recognition systems. Pervasive Mob Comput 3:300–327CrossRefGoogle Scholar
  29. 29.
    Sun FT, Kuo C, Griss M (2011) Pear: Power efficiency through activity recognition (for ecg-based sensing). In: Pervasive computing technologies for healthcare (PervasiveHealth), 2011 5th International conference, pp 115–122Google Scholar
  30. 30.
    Thatte G, Li M, Lee S, Emken A, Narayanan S, Mitra U, Spruijt-Metz D, Annavaram M (2012) Knowme: an energy-efficient multimodal body area network for physical activity monitoring. ACM Trans Embed Comput Syst 11(S2):48:1–48:24CrossRefGoogle Scholar
  31. 31.
    Van Laerhoven K, Kilian D, Schiele B (2008) Using rhythm awareness in long-term activity recognition. In: Proceedings of the 12th international symposium on wearable computers (ISWC 2008). IEEE Press, Pittsburgh, pp 63–68Google Scholar
  32. 32.
    Wang Y, Krishnamachari B, Annavaram M (2012) Semi-Markov state estimation and policy optimization for energy efficient mobile sensing. In: The 9th annual IEEE communications society conference on sensor, mesh and Ad Hoc communications and networks (SECON’12)Google Scholar
  33. 33.
    Wood A, Merrett G, Gunn S (2012) Adaptive sampling in context-aware systems: a machine learning approach. IET wireless sensor systems 2012Google Scholar
  34. 34.
    Yan Z, Subbaraju V, Chakraborty D, Misra A, Aberer K (2012) Energy-efficient continuous activity recognition on mobile phones: an activity-adaptive approach. In: 16th international symposium on wearable computers, pp 17–24Google Scholar

Copyright information

© Springer-Verlag London 2013

Authors and Affiliations

  1. 1.Karlsruhe Institute of Technology (KIT)KarlsruheGermany

Personalised recommendations