Citation
Armin Wasicek. "Composing Learning Internet-of-Things Applications with Accessors". Talk or presentation, 16, October, 2015; Presented at the Eleventh Biennial Ptolemy Miniconference, Berkeley.

Abstract
Learning Internet-of-Things (IoT) applications have three ingredients: Physical components that provide a useful service; for example in a car this would be the engine or the wheels. Smart components comprise elements that increase the value of the physical components, like sensors, microprocessors, data storage, controls, and software; smart components do not only collect information, but also organizing and interpreting that information; an automotive example is Adaptive Cruise Control (ACC) that intelligently adapts a car's set speed to match that of the car in front. Finally, connectivity enables smart components and thereby physical components to participate in larger aggregations, thus, forming a "swarm" of things ; swarms can gather and disband, and smart components can be part of several swarms.

"Accessors" are a design pattern to build composable IoT applications, so-called "swarmlets". The intent of accessors is to enable interoperability between the myriad of frameworks and services that enable the IoT. Accessors are wrappers for sensors, actuators, and services that export an actor interface. An actor semantics provides ways to compose accessors with disciplined and understandable concurrency models. We argue that accessors facilitate building smart systems tremendously by providing streaming interfaces to sensors, actuators, and network services. For that purpose, accessors embody horizontal and vertical design contracts. The horizontal design contract is that of Actor composition, and it enables an accessor to interact with other accessors and actors at the same level of abstraction. The vertical design contract is connection between the underlying network service, such as a internet API or a device driver, and the actor-oriented interface exposed in the design.

Ptolemy II implements an experimental swarmlet host that facilitates integrating different paradigms by agree.ment on fundamental design decisions like data representation, interface design, protocol selection, and execution semantics. For instance, Ptolemy II provides a Discrete Event (DE) director that can be used as a baseline Model of Computation (MoC) for the composition of accessors. It provides PILOT, a machine learning toolkit that is particularly designed to work with streaming interfaces. And its lattice-based ontology package enables ontological inference, reducing the need for explicit type annotation for programmers.

Electronic downloads

• Wasicek_ComposingLearningIoTApplicationsWithAccessors_PtolemyMiniConference_16Oct2015.pdf · application/pdf · 1757 kbytes

• HTML

  Armin Wasicek. <a
  href="http://chess.eecs.berkeley.edu/pubs/1125.html"><i>Composing
  Learning Internet-of-Things Applications with
  Accessors</i></a>, Talk or presentation,  16,
  October, 2015; Presented at the <a
  href="http://ptolemy.eecs.berkeley.edu/conferences/15/"
  >Eleventh Biennial Ptolemy Miniconference</a>,
  Berkeley.

• Plain text

  Armin Wasicek. "Composing Learning Internet-of-Things
  Applications with Accessors". Talk or presentation, 
  16, October, 2015; Presented at the <a
  href="http://ptolemy.eecs.berkeley.edu/conferences/15/"
  >Eleventh Biennial Ptolemy Miniconference</a>,
  Berkeley.

• BibTeX

  @presentation{Wasicek15_ComposingLearningInternetofThingsApplicationsWithAccessors,
      author = {Armin Wasicek},
      title = {Composing Learning Internet-of-Things Applications
                with Accessors},
      day = {16},
      month = {October},
      year = {2015},
      note = {Presented at the <a
                href="http://ptolemy.eecs.berkeley.edu/conferences/15/"
                >Eleventh Biennial Ptolemy Miniconference</a>,
                Berkeley},
      abstract = {Learning Internet-of-Things (IoT) applications
                have three ingredients: Physical components that
                provide a useful service; for example in a car
                this would be the engine or the wheels. Smart
                components comprise elements that increase the
                value of the physical components, like sensors,
                microprocessors, data storage, controls, and
                software; smart components do not only collect
                information, but also organizing and interpreting
                that information; an automotive example is
                Adaptive Cruise Control (ACC) that intelligently
                adapts a car's set speed to match that of the car
                in front. Finally, connectivity enables smart
                components and thereby physical components to
                participate in larger aggregations, thus, forming
                a "swarm" of things ; swarms can gather and
                disband, and smart components can be part of
                several swarms. <p>"Accessors" are a design
                pattern to build composable IoT applications,
                so-called "swarmlets". The intent of accessors is
                to enable interoperability between the myriad of
                frameworks and services that enable the IoT.
                Accessors are wrappers for sensors, actuators, and
                services that export an actor interface. An actor
                semantics provides ways to compose accessors with
                disciplined and understandable concurrency models.
                We argue that accessors facilitate building smart
                systems tremendously by providing streaming
                interfaces to sensors, actuators, and network
                services. For that purpose, accessors embody
                horizontal and vertical design contracts. The
                horizontal design contract is that of Actor
                composition, and it enables an accessor to
                interact with other accessors and actors at the
                same level of abstraction. The vertical design
                contract is connection between the underlying
                network service, such as a internet API or a
                device driver, and the actor-oriented interface
                exposed in the design. <p>Ptolemy II implements an
                experimental swarmlet host that facilitates
                integrating different paradigms by agree.ment on
                fundamental design decisions like data
                representation, interface design, protocol
                selection, and execution semantics. For instance,
                Ptolemy II provides a Discrete Event (DE) director
                that can be used as a baseline Model of
                Computation (MoC) for the composition of
                accessors. It provides PILOT, a machine learning
                toolkit that is particularly designed to work with
                streaming interfaces. And its lattice-based
                ontology package enables ontological inference,
                reducing the need for explicit type annotation for
                programmers. },
      URL = {http://chess.eecs.berkeley.edu/pubs/1125.html}
  }
  

Posted by Christopher Brooks on 19 Oct 2015.
Groups: ptolemy
For additional information, see the Publications FAQ or contact webmaster at chess eecs berkeley edu.