Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision		 Previous revision
Last revisionBoth sides next revision
lirec_notes [2009-01-06 15:08] – davegriffithslirec_notes [2020-06-05 22:11] – old revision restored (2009-01-08 10:33) 114.119.160.215
Line 1: Line 1:
 ==== LIving with Robots and InteractivE Companions ==== ==== LIving with Robots and InteractivE Companions ====
  
-LIREC aims to establish a multi-faceted theory of artificial long-term companions (including memory, emotions, cognition, communication, learning, etc.), embody this theory in robust and innovative technology and experimentally verify both the theory and technology in real social environments. Whether as robots, social toys or graphical and mobile synthetic characters, interactive and sociable technology is advancing rapidly. However, the social, psychological and cognitive foundations and consequences of such technological artefacts entering our daily lives - at work, or in the home - are less well understood.+<blockquote>LIREC aims to establish a multi-faceted theory of artificial long-term companions (including memory, emotions, cognition, communication, learning, etc.), embody this theory in robust and innovative technology and experimentally verify both the theory and technology in real social environments. Whether as robots, social toys or graphical and mobile synthetic characters, interactive and sociable technology is advancing rapidly. However, the social, psychological and cognitive foundations and consequences of such technological artefacts entering our daily lives - at work, or in the home - are less well understood.</blockquote>
  
-==Agent platforms, and migration between them==  +==Links== 
-An interesting feature of agents which need to build up a long term relationship with their users will moving between different platforms. The platforms currently under investigation are:+ 
 +http://www.lirec.org 
 + 
 +[[lirec work packages]] 
 + 
 +==Agent platforms== 
 + 
 +The platforms currently under consideration for long term companions are:
   * Mobile robots   * Mobile robots
   * Fixed robots   * Fixed robots
Line 10: Line 17:
   * Fixed graphical systems   * Fixed graphical systems
  
-This migration between these devices, and how people relate to it, is a core element of the research.+==Migration== 
 +  
 +An interesting feature of the research is migration between these platforms. Agents which need to build up a long term relationship with their users will have to switch forms depending on the needs of the user at different times. The migration of an agent between devices, and how people relate to it, is a core element of the research.
  
 ===Scenarios=== ===Scenarios===
  
-The scenarios are to be considered preliminary ideas for showcasing the technology developed for lirec, and highlight companionship and mostly exercise migration between devices.+In order to test and showcase the technology developed for Lirec, several scenarios have been designed to promote companionship. These scenarios are shared between three of the research partners.
  
 ==Heriot Watt== ==Heriot Watt==
 Spirit of the building: Spirit of the building:
-  - Team buddy, mobile robot, collective memory for a lab team +  - Team buddy, a mobile robot/collective memory for a team working in a lab 
-  - Personal guide - for navigating around a university campus, remember appointments etc +  - Personal guide - for navigating around a university campus, remembering appointments, telling you where to go 
-  - In the wild - a gossip/chat robot - appears on a large screen in social area+  - In the wild - a gossip/chat robot - appears on a large screen in a social area
  
 ==INESC-ID== ==INESC-ID==
   - Game companion for young children   - Game companion for young children
-  - Personal trainer (migrate to mobile robot for jogging exercises etc) +  - Personal trainer - which can migrate to mobile robot for jogging exercises 
-  - Welcome to the jungle - talk to game characters through a robot, robot can alternate  +  - Welcome to the jungle - talk to game characters through a robot which can alternate between real and game world
-    between real and game world+
  
-==University of Hertfordshire== +==The University of Hertfordshire== 
-  - Fetch and carry, help with physical impairment or convenience +  - Fetch and carry, to help with physical impairment or provide convenience 
-  - Cognitive prosthetic - memory aid for tasks etc +  - Cognitive prosthetic - memory aid for tasks 
-  - Telepresence card player - robot mediates play +  - Telepresence card player - a robot mediates play between two people 
-  - Teaching Proxemic preferences - robot learns where to be relative to the user in different situations +  - Teaching proxemic preferences - a robot learns where to be relative to the user in different situations 
-  - Travelling companion - agent migration, to stay with user during home, work, shopping.+  - Travelling companion - agent migration, to stay with user during home, work, shopping 
 + 
 +===Experimental testing=== 
 + 
 +All scenarios developed are to be experimentally tested and showcased to the public.
  
 ===Architecture=== ===Architecture===
  
-Has to:+The technology developed for Lirec is shared between the research partners, and has to:
   - Run on very different platforms   - Run on very different platforms
   - Reuse code across these platforms   - Reuse code across these platforms
   - Support migration at runtime between platforms   - Support migration at runtime between platforms
- 
-Platforms can consist of 4 main types, mobile robot, fixed robot, handheld device or fixed graphical system. Each has it's inherent restrictions.  
  
 ==Existing robot architectures== ==Existing robot architectures==
-In the whole, there is a lack of sharing of this kind of technology. This is partly because generalising is hard in this field, considering all types of robots possible. However, Lirec has to generalise as it's using a wide variety of architectures.+In the whole, there has historically been a lack of sharing of this kind of technology. This is partly because generalising is hard, considering all types of robots and implementations possible. However, Lirec has to generalise as it's using a wide variety of platforms and needs to share as much as possible.
  
-NASREM/NIST RCS - NASA + ESA use a generic system with their subcontractors.+While some existing architectures are available, none of them address the needs of Lirec, given the focus is: 
 +  - Long term companionship rather than solving a purely technical problem 
 +  - Working across different platforms
  
 ==Methodology== ==Methodology==
-In the past there have been 2 broad approaches to robot design: +In the past there have been 2 broad approaches to [[robot design]]: 
-  * Hierarchical, model based planning = expensive to maintain accurate world state model +  * Hierarchical, model based planning = expensive/difficult to maintain an accurate world state model 
-  * Behavioural approach = less state, local decisions, liable to local minima, opaque to program +  * Behavioural approach = less state, local decisions, but liable to local minima and opaque to program 
-This can be summed up as predictive vs reactive.+This can be summed up as "predictive vs reactive".
  
-The plan is to use a hybrid approach, example BIRON. Where the predictive constrains the reactive to combine local decisions with a world model.+The plan is to use a hybrid approach, example [[http://www.techfak.net/ags/ai/projects/BIRON/welcome.html|BIRON]]. Where the high level predictive element constrains the reactive in order to combine local decisions with a world model.
  
 The architecture will consist of 3 layers of abstraction: The architecture will consist of 3 layers of abstraction:
-  * Level 1 - device layer, platform dependant+  * Level 1 - The device layer, hardware drivers
   * Level 2 - platform dependant -> logical mappings   * Level 2 - platform dependant -> logical mappings
   * Level 3 - ION, platform independent   * Level 3 - ION, platform independent
Line 64: Line 76:
 Level 2 will provide a reference architecture with modular capabilities called competencies. Not all competencies will make sense for all platforms, and different implementations of the same competency will be needed for different platforms. Level 2 will provide a reference architecture with modular capabilities called competencies. Not all competencies will make sense for all platforms, and different implementations of the same competency will be needed for different platforms.
  
-Example competencies: +Competencies table 
-  * Face finding +^ Actuation ^^^^ Sensing ^^^^^^ 
-  * Expression recognition +| **Speech/Sound** | **Visual** | **Movement** | **Object Manipulation** | **Identification** | **Vision** | **Sounds** | **Positioning** | **Distance** | **Internal State** | 
-  * Text to speech +| Text to speech | Gesture execution | Move limb | Grasp/Place object | User recognition | Face detection | Speech recogn | Localization | Obstacle avoidance | Battery status | 
-  * Obstacle avoidance+| Non-verbal sounds | Lip sync | Follow person | | Obj recognition | Gesture recognition (simple set) | Non-verbal sounds | Locate person | Locate obj | Competence execution monitoring | 
 +| | Facial expression | Path planning | | | Emotion recognition (simple set) | | | User proximic distance sensing/control | | 
 +| | | Gaze/head movement | | | Body tracking | | | | | 
 +| | | Expressive behaviour | | | | | | | | 
 + 
 +The computer vision competencies are to be concentrated on at first. 
 + 
 +[[http://eris.liralab.it/yarp/|YARP]] is a good example of sharing code in this way on a Linux platform.
  
-YARP 
  • lirec_notes.txt
  • Last modified: 2020-06-05 22:30
  • by nik
  • Currently locked by: 47.128.63.52