Applications of and Challenges in Self-Awareness

All participants of the Slides Factory

Application 1: SwarmRobotics

• Imagine a swarm of robots that need to solve a certain task, e.g.– Cleaning a devastated area– Exploring Mars

• In difficult environments with holes, hills, obstacles, . . . the robots have to cooperate– Transport an object together– Form organisms to cope better

with environment

Application 1: SwarmRobotics

• Robots are aware of the task they are supposed to perform and monitor their performance in the environment

• Robots should be able to adapt to maximize their performance

• Adaptations take place on an individual level as well as on a collective level:

– Individuals adjust their behavior

– Collective behavior emerges (e.g. organisms are formed by multiple robots)

Example project – SYMBRION (1)

Symbiotic Evolutionary Robot Organisms

• Hundreds of small cubic robots are built and deployed in an environment

• Robots sense each other and the environment and are capable of aggregating into “multi-cellular” organisms

• Aggregation and disaggregation is self-driven, depending on the circumstances: different environments, different tasks

• Questions addressed:– Can we build such robots and program the basic behaviors needed for

appropriate (dis)aggregation?

– Can we provide adaptive mechanisms that enable newly “born” organisms learn to operate (sense, move, act, …)?

Example project – SYMBRION (2)

Scenario movie

http://www.youtube.com/watch?v=SkvpEfAPXn4

Example project – SYMBRION (3)

Approach

Example project – SYMBRION (4)

Current Results

• Different controllers have been developed for robots

• Evolutionary approaches are able to adapt the controllers based upon fitness

• Different organisms are formed as required by the environment

• Some initial versions of hardware have been developed and are currently being deployed

Example project – ASCENS (1)

Autonomous service component ensembles

• Self-aware, self-adaptive, and self-expressive autonomous components

• Components run in an environment and are called ensembles

• Systems are very difficult to develop, deploy, and manage

• Goal of ASCENS: – Develop an approach that combines traditional SE approaches based

on formal methods with the flexibility of resources promised by autonomic, adaptive, and self-aware systems

• Case studies:– Robotics, cloud computing, and energy saving e-mobility

Example project – ASCENS (2)

Approach

Example project – CoCoRo (1)

Collective Cognitive Robotics

• Aims at creating an autonomous swarm of interacting, cognitive underwater vehicles

• Tasks to be performed by the swarm:– Ecological monitoring

– Searching

– Maintaining

– Exploring

– Harvesting resources

Example project – CoCoRo (2)

Scenario movie

http://www.youtube.com/watch?v=OStLml7pHZY

Example project – CoCoRo (3)

Approach

• Draw inspiration from nature to generate behavior:– Cognition generating algorithms:

• Social insect trophallaxis

• Social insect communication

• Slime mold

• ANN

– Collective movement:• Bird movement

• Fish school behavior

Application 2: Power networks

• Current power networks rely mainly on big companies, generating and distributing energy

• The scenario is quickly changing:– Renewable energy (solar panels, wind turbines, …)

– “Home-made” energy

– Smart devices

• This opens to a lot ofopportunities, butrequires an appropriatemanagement

A new scenario

• People can produce their own energy

• People can sell energy they do not use

– To their neighbors in a peer-to-peer fashion

• Renewable energy impacts positively on the environment

• Smart devices can help in controlling the energy consumption and in providing us with information

Renewable

• US Nationwide energy dispatch without (a) and with (b) renewable contribution

• Source: Brinkman, Denholm, Drury, Margolis, and Mowers, “Toward a solar- powered grid,” Power and Energy Magazine, IEEE, vol. 9, no. 3, pp. 24–32, 2011

The new scenario’s issues

• The new scenario introduces some peculiarities– The production is “distributed” among a possibly large

number of producers (or “prosumers” if they consume energy)

– The production is subject to external conditions (e.g., weather)

– Smart devices are better than old ones but must be coordinated

• In general, we have a more dynamic and unpredictable scenario

Power network control

• But how this situation can be controlled?

• A human control

– Is difficult (many parameters, autonomous entities, …)

– Can be not impartial (big companies are self-interested)

• Can a power network control itself?

What is needed?

• In both cases, for networks’ self management/organization we need:

– Mechanisms, which can enable the network to act on itself

– Policies or goals, which leads the networks in taking decisions

Example project - PowerTAC

• Represent each house by means of an agent

• Agents are aware of their current and expected future energy expenditure

• Agents act based upon this knowledge

• Can either sell or buy energy

• PowerTAC: competition to develop appropriate mechanisms and agents for selling and buying energy

Application 3: Data management

• More and more content is being generated

• Content needs to be effectively managed in order to avoid user form being swamped

• Task is to:

– Manage existing content

– Acquire new content

Example project - SAPERE

Self-aware Pervasive Service Ecosystems

• Computers for handling data and providing services are integrated into an “ecosystem”

• System is extended with – methods for data and situation identification

– decentralized algorithms for spatial self-organization, self-composition, and self-management

• Thus, we obtain automated deployment and execution of services and for the management of contextual data items

Scenario

• Pervasive computing– Sensor rich and always connected smart phones– Sensor networks and information tags– Localization and activity recognition– Internet of things and the real‐time Web

• Innovative pervasive services arising– Situation‐aware adaptation– Interactive reality– Pervasive collective intelligence and pervasive participation

• Open co‐production scenario, very dynamic, diverse needs and diverse services, continuously evolving

Architecture

• Open production model• Smooth data/services

distinction– live semantic annotations (LSA)

• Interactions– Sorts of bio‐chemical reactions

among components– In a spatial substrate

• Eco‐laws– Rule all interactions– Discovery + orchestration

seamlessly merged

• Built over a pervasive network world

Infrastructure and applications

• Infrastructure– A very lightweight infrastructure– Ruling all interactions (from discovery to data exchange and

synchronization) by embedding the concept of eco‐laws– To most extent, acting as a recommendation and planning engine– Possibly inspired by tuple space coordination models– Yet made it more “fluid” and suitable for a pervasive computing

continuum substrate not a network but a continuum of tuple spaces

• Applications– The “Ecosystem of Display” as a general and impactfultestbed– To put at work and demonstrate the SAPERE findings– Active and dynamic information sharing in urban scenarios– Active participation of citizens to the working of the urban

infrastructure

Example project - RECOGNITION

Relevance and Cognition for Self‐Awareness in a Content‐Centric Internet• Project draws inspiration from human cognitive processes to

achieve self-awareness• Try to replicate core cognitive processes in computer systems:

– e.g. inference, beliefs, similarity, and trust– embed them in ICT

• Application domain: internet content– Manage and acquire content in an effective manner by means of

self-aware computing systems

Motivation: Technological Trends

• Participatory generation of content– Prosumers, diversity, expanding edges– Long tail, swamping, scale!

• Content in the environment– Linkage of the physical and virtual worlds– Embedding content and knowledge

• Acquiring knowledge through social mechanisms– Blogging, social networking, recommendation, RSS

feeds…

• How content reaches users will continue to change…

Self-awareness to support technological trends

• Intention: Paradigm to support ICT functions

– Enabling content centricity

• Better fitting of users to content and vice-versa

– Synchronize content with human activity and needs

• Place, time, situation, relevance, context, social search

– Autonomic management

• Of content, its acquisition and resource utilization

Approach: Human Awareness Behaviour

• Capture & exploit key behaviours of the most intelligent living species– Human capability is phenomenal in navigating

complex & diverse stimuli

– Filter & suppress information in “noisy” situations with ambient stimuli

– Extract knowledge in presence of uncertainty

– Exercise rapid value judgment for prioritisation

– Engage a and multi‐scale social context multi learning

Application 4: Cooperative E-Vehicles

• In a few years the e-mobile cars of a big town will be able to communicate with• each other and the time tables of the users

• traffic management servers,

• battery loading stations,

• parking lots, etc.

• In such an ensemble, the communicating entities and users may pursue different goals and plans– several users may share cars, but have different time tables

– Loading stations have only limited capabilities; so cars may not be able to use the nearest station for changing the battery

Application 4: Cooperative E-Vehicles

• Communication and cooperation between the entities of the ensemble leads to better Quality of Service w.r.t.

– reliability

• e.g. transport/delivery reliability, adherence to schedules, guarantee to reach the goal, recharging-in-time assurance

– adaptability to changes

• e.g. traffic flow, daily personal schedule of the driver

– predictability of plans

• confidence in reaching a desired location at a preferred time

Application 5: Science Cloud

• consists of a collection of notebooks, desktops, servers, or virtual machines– running a cloud platform

/application– communicating over the Internet

(IP protocol), forming a cloud– providing data storage and

distributed application execution • Every participant is

– provider and possible user of resources

– knowsabout• itself(properties set by

developers),• its infrastructure (CPU load,

available memory),and• other SCPis(acquired through

the network)

Application 5: Science Cloud

• The science cloud – is dynamically changing

• Participants may dynamically join or leave the cloud or just disappear from the cloud

– is fail-safe • Continues working if one or several nodes fail

– provides load balancing • By parallelly executing applications if the load is high, but

not before that.

– aims at energy conservation• virtual machines are shut down or are taken out of the

configuration if not required

Current research questions and challenges

• Dilemma of wishing to make our designed artefacts autonomous but not too much (safety).

• To have a metrics to measure properties related to awareness, autonomy.

• We do not know how to engineer self-organization and emergence.

• We do not know how to cope with autonomy and variability. Dilemma of system stability and reliability incorporating randomness and variability.

• How to design and implement self-aware systems?

• What kind of tools and methodology can we use here?

• Is it ethical to build self-aware systems?

• Can we build autonomic self-aware systems that behave in an ethical way? Related: legally correct behaviour, behaviour compliant with some set of rules and regulations.

• What makes known natural systems self-aware?

• Describing the scope of the future behaviour of a self-aware system.

Current research questions and challenges

• Predicting the behaviour of autonomic systems and their interactions with the environment.

• How to ensure safety and security of autonomic self-aware systems? How to differentiate malicious from benign behaviour?

• What does the system theory of autonomic self-aware systems look like?

• How to build an autonomic self-aware system that would last 100 years?

• To what extent can Big Data be treated as an autonomic self-aware system?

• Can you separate an autonomic self-aware system from its environment?

• In what sense is human and machine self-awareness different? What implications do these differences have on developing them?

• How can we draw inspiration from human self-awareness for designing machine self-awareness?

• How to do the second order design needed in autonomic self-aware systems?

• Will autonomic self-aware systems develop their own medical science?

• Goal: build an autonomic self-aware energy production system.

• Goal: build a smart city / computer network / communication network.

References

• Sapere

– http://www.sapere-project.eu/

– C. Villalba and F. Zambonelli, "Towards Nature-Inspired Pervasive Service Ecosystems: Concepts and Simulation Experiences", Journal of Network Computers and Applications, vol. 34(2), pp.589-602

– F. Zambonelli, "Pervasive Urban Crowsourcing: Visions and Challenges", The 7th IEEE Workshop on PervasivELearning, Life, and Leisure (PerEl 2011), pp.578-583, 21-25 March 2011