Ubiquitous Network Robots for Life Support

Transcription

1 DAY 2: EXPERTS WORKSHOP Active and Healthy Ageing: Research and Innovation Responses from Europe and Japan Success Stories in ICT/Information Society Research for Active and Healthy Ageing Ubiquitous Network Robots for Life Support October 10, 2012 Dr. Takahiro Miyashita Group Leader, Dept. of Network Robots ATR Intelligent Robotics and Communication Labs This research is supported by the Ministry of Internal Affairs and Communications.

2 OUTLINE 1. Robotic Services 2. Ubiquitous Network Robots (UNR) for Life Support 3. UNR Platform and its Standardization 4. Conclusion

3 1. Robotic Services

4 Robotic Services for Active & Healthy Ageing Shopping support Active hearing Health care Touring support Community building support

5 However, Stand-alone robotic services are not enough for continuously and seamlessly supporting daily activity.

6 Real World in Social Interaction with or via Robots (2002) Noisy, Noisy, Noisy Noisy,

7 Networked Robots in School (2002) wireless tag Antenna

8 Structuring Environmental Information on Human Trajectories and Behaviors (UbiComp 2008) T. Kanda et al., Who will be the customer?: A social robot that anticipates people s behavior from their trajectories, Proc. of UbiComp 2008, pp , 2008.

9 Ambient Intelligence Map Anticipation of People s Trajectories and Behaviors T. Kanda et al., Abstracting People s Trajectories for Social Robots to Proactively Approach Customers, IEEE Trans. on Robotics, Vol.25, No.6, pp , 2009.

10 Robots in Smartphone (2010) CG character agent He can move and talk in a display with same commands for robots in real world.

11 Three Types of Robots with UNR Platform ICT Infrastructure (Cloud and Ubiquitous Computing & Wireless Communication,) Visible-type Physical Robot, Geminoid, WheelChair User Interface: Broaden to Elderly Pricecompetitive to bike or vehicle, (or hobby) Virtual-type iphone, ipad, DS-3D, User Interface easy-to-use & inevitable utility in daily use Price: Reasonable Ubiquitous Network Robot(UNR) Platform Unconscious-type RF-ID, Camera, LRF, Location Precision: 3m5cm accuracy Pricegradually decreasing

12 2. Ubiquitous Network Robots (UNR) for Life Support

13 UNR Research Project (Supported by MIC) Purpose - Create core technologies allowing network robot services across multiple areas for elderly and disables (challenged) which are hardly realized by single robots. - Propose the common platform, called the Ubiquitous Network Robot PlatForm (UNR-PF), which provides the common functionalities on the services of human-robot interaction. Period Budget Partners June, 2009 to March, M US$ ATR, Toshiba, NTT, Hitachi, and NEC

14 Ubiquitous Network Robots for Life Support Human-Robot Interaction One location to Multi-location Ubiquitous Network Robots Platform -R Collaboration and Coordination Tele-operated Communication

15 Shopping Support for Elderly Customer (2009) Teleoperator Home Shopping Area Entrance of Super Market

16 Networked Robots help elderly customers in a supermarket (December, 2009)

17 Shopping Support with UNR over three area AreaHome Area Entrance of a Supermarket AreaShopping Area 17

18 Networked Robots help elderly customers in a supermarket (December, 2009) Comments from Elderly Customer Participant

19 Shopping support service for elderly wheelchair users

20 Shopping support service for elderly wheelchair users (March, 2011)

21 3. UNR Platform and its Standardization

22 Conventional Style of Service Application Programming Service App. 1 find face guest name face position move to the position Service App. 2 find tag tag ID tag position walk to the position Implementation-dependent functions called from service applications and their return values face recognition wheel control RFID tag Detection leg control Implementation layer (hardware layer) robot 1 robot 2 Application program and robot implementation are tightly coupled

23 Proposed Service Application Programming Style Who is there? Service App. X person ID position approach the person compatible Standard Interfaces and Framework Who is there? Service App. X person ID position approach the person Implementation -independent functions and their return values in common message type face recognition wheel control RFID tag Detection leg control Implementations are invisible from outside HRI Engine 1 HRI Engine 2 robot 1 robot 2 Standard APIs enhance the reusability of application programs

24 Proposed Service Application Programming Style Service App. Service X compatible Apps. Service App. X Who is there? person ID position face recognition HRI Engine 1 approach the person wheel control Standard Interfaces and Framework Who is there? person ID position RFID tag Detection leg control HRI Engine 2 approach the person UNR-PF Layer Robotic Compos. Implementation -independent functions and their return values in common message type Implementations are invisible from outside robot 1 robot 2 Standard APIs enhance the reusability of application programs

25 Ubiquitous Network Robot Platform separates the developments of robots and robotic service applications. connects various service applications to appropriate robots in a variety of places. permits service applications to share their own data between each other. UNR-PF alpha (test release) is available. K. Kamei et al., Cloud Networked Robotics, in IEEE Network Magazine, vol. 26, no. 3, pp , May

26 Implementation Model of UNR Platform Robotic Service Robotic Service Management (Global (multi-area) Integration) Robot Registry (global) Local Platform (Location B) Platform Service Registry (global) Resource Management Map Registry Local Platform (Location A) Command Service Registry Network Robot Interface User Registry Status: 4W =Who/When/ Where/ What Component Component Component Component Component Component Component

27 Standardizations Robotic Service Robotic Service Management (Global (multi-area) Integration) Robot Registry (global) Local Platform (Location B) Platform CityGML Service Registry (global) Resource Management Map Registry RLS Local Platform (Location A) UNR-PF Command Service Registry Network Robot Interface User Registry Status: 4W =Who/When/ Where/ What Component Component Component RoIS Component Component Component Component

28 Standardization Map Sensor Network Indoor Map OGC GIS ISO/TC 211 Industrial Robots ISO/TC 184 IEC/TC61 Service Robots Electrical Appliances Distributed System OMG Service Robots Device IEEE Network ITU-T SG16 Multimedia Sensor Network

29 Standardization Map GML Sensor Network Indoor Map GeoSMS OGC Robotics components RTC DDC4RTC OMG Distributed System CityGML Service Robots ISO191XX series IndoorGML Robotic Interaction Service RoIS Robotic Localization Service RLS Mapdata GIS ISO/TC 211 Device IEEE Glossary/ontology Industrial Robots Vocabulary ISO/TC 184 PersonalCare IEC/TC61 Safety Life Cycle ORiN Industry Safety Care Safety Safety of households Network Robot ITU-T SG16 Platform Network Service Robots Electrical Appliances Multimedia Sensor Network

30 Future work for Creating Robotic Services Market Scientific and Technological Challenges System Reliability and Standardization Insurance and legal issues Collaboration with local administrations Design service & Evaluation tests User Acceptance End-User Analysis & Education Cross fertilization: academics and industry Cost Performance

31 Conclusion The concept of Ubiquitous Network Robot and some of robotic service applications based on UNR were introduced. The UNR is one of key concepts for supporting Active and Healthy Ageing. Let s collaborate with not only robots but also us (!) to develop the robotic services! Thank you for your kind attention. This research is supported by the Ministry of Internal Affairs and Communications.