T : Internet Technologies for Mobile Computing

Transcription

1 T : Internet Technologies for Mobile Computing Overview of IoT Platforms Julien Mineraud Post-doctoral researcher University of Helsinki, Finland Wednesday, the 9th of March 2016 Julien Mineraud Overview of IoT Platforms / 37

2 Outline 1 Introduction 2 Evaluation of the IoT platform landscape 3 Gap analysis of IoT plaforms 4 Useful IoT platforms 5 IoT Hub & IoT Market Julien Mineraud Overview of IoT Platforms / 37

3 Introduction Outline 1 Introduction 2 Evaluation of the IoT platform landscape 3 Gap analysis of IoT plaforms 4 Useful IoT platforms 5 IoT Hub & IoT Market Julien Mineraud Overview of IoT Platforms / 37

4 Introduction IoT platforms Julien Mineraud Overview of IoT Platforms / 37

5 Introduction Overview IoT platforms Nowadays, many Internet-of-Things (IoT) platforms are available to ease the manipulation of smart objects, such as wireless controlled lights, mobile devices and sensors. These platforms are often different in nature, because they are designed to solve different problems: Type Home automation Smart factory Architecture Centralized Cloud-based Nb. of devices Small (up to 500) Large Storage Local Cloud Focus User-friendly Predictive analytics Julien Mineraud Overview of IoT Platforms / 37

6 Evaluation of the IoT platform landscape Outline 1 Introduction 2 Evaluation of the IoT platform landscape 3 Gap analysis of IoT plaforms 4 Useful IoT platforms 5 IoT Hub & IoT Market Julien Mineraud Overview of IoT Platforms / 37

7 Evaluation of the IoT platform landscape Criteria for evaluating IoT platforms Support of heterogeneous devices Type Architecture Openness of the project Availability of Application Programming Interface (API) Accessibility of the data and devices Discoverability of devices, data and other services Julien Mineraud Overview of IoT Platforms / 37

8 Evaluation of the IoT platform landscape Heterogeneity of devices Julien Mineraud Overview of IoT Platforms / 37

9 Evaluation of the IoT platform landscape Type of platforms & architecture Type PaaS: Platform-as-a-Service Refers to the platforms that provide cloud computing services for IoT devices and data. The services include, but are not restricted to storage facilities, devices management, device connectivity, backup mechanisms or online support SaaS: Software-as-a-Service SaaS focuses on the mashup of data using cloud computing capabilities Local server or hub: Unliketheprevioustwo,thistypeofIoT platforms are not provisioned in the cloud Architecture Can be either cloud-based, centralized, decentralized or a combination of centralized or cloud-based Julien Mineraud Overview of IoT Platforms / 37

10 Other criteria Evaluation of the IoT platform landscape Openness Most IoT platforms are proprietary, but many good open-source IoT platforms are available (mostly for home context) API Almost all platforms provide a REST API GET, POST, PUT, DELETE Discoverability Only a few platforms have integrated some type of service discovery mechanisms, even in a very simplified fashion Julien Mineraud Overview of IoT Platforms / 37

11 Evaluation of the IoT platform landscape Security One of the fundamental criteria for IoT platforms is the need to include efficient and reliable privacy and security mechanisms Cloud-based IoT platforms are prone to traditional web and network security attacks such as Denial of Service (DoS), man-in-the-middle, eavesdropping, spoofing and controlling attacks AnumberofareasarecriticalforthewidespreadadoptionofIoTbut not yet fully addressed by IoT platforms: device authentication communication and physical privacy data storage protection device protection trust management fault tolerance Julien Mineraud Overview of IoT Platforms / 37

12 Evaluation of the IoT platform landscape Evaluation of a few IoT platforms Full evaluation at Julien Mineraud Overview of IoT Platforms / 37

13 Evaluation of the IoT platform landscape Interactions between IoT platforms Julien Mineraud Overview of IoT Platforms / 37

14 Gap analysis of IoT plaforms Outline 1 Introduction 2 Evaluation of the IoT platform landscape 3 Gap analysis of IoT plaforms 4 Useful IoT platforms 5 IoT Hub & IoT Market Julien Mineraud Overview of IoT Platforms / 37

15 Gap analysis of IoT plaforms Gap analysis of IoT platforms Support of heterogeneous devices Data ownership Data processing & sharing Developer support Ecosystem formation IoT marketplace Julien Mineraud Overview of IoT Platforms / 37

16 Gap analysis of IoT plaforms Support of heterogeneous devices Current status: Platforms assume smart objects to talk HTTP or require gateway Expectations: Devices must be easily and securely integrable to the IoT platform without a gateway Unified resources and simplify usability Gaps: Support of constrained devices Standardized IoT devices models Secure authentication, identification and management of IoT devices Problems: Heterogeneous interactions Protocol standardization Recommendations: Relying on standard protocols (e.g., CoAP, LwM2M, MQTT) Integration of state-of-the-art security and privacy protocols Julien Mineraud Overview of IoT Platforms / 37

17 Data ownership issue Gap analysis of IoT plaforms Current status: Mainly given to the end-user but with very simple privacy policies Expectations: Full control given to the owner of the data Local storage Fine-grained data visibility model Gaps: Manipulation of data in edge devices Self-storage Problems: Security of the data storage Device constrains to store data and provide secure access control Recommendations: Algorithms and mechanisms available to the data owner to limit the access only to a predefined set of the resources Julien Mineraud Overview of IoT Platforms / 37

18 Gap analysis of IoT plaforms Data processing & sharing Current status: Nonuniform data sharing format Sharing is performed via nonuniform REST API Expectations: Uniform data format across multiple platforms Pub/Sub mechanism and data catalogs Edge analytics Gaps: Data processing is not well integrated in IoT platforms Data analytics is only available in cloud-based solutions Data catalogs are missing Problems: Fusion efficiently data streams from multiple data catalogs IoT devices have limited computing capabilities Recommendations: Data catalogs with semantic indexes Uniform and interoperable data models Integration of data processing technologies in platforms Julien Mineraud Overview of IoT Platforms / 37

19 Developer support Gap analysis of IoT plaforms Current status: REST API to access the data or devices handled by the platform Applications are for internal use rather than for sharing (except IFTTT) Expectations: Common API to ease the development of cross-platform applications Domain Specific Language (DSL) dedicated to cross-platform application development Gaps: Application mashup APIs Limited presence of SDKs Absence of DSL with higher abstraction level primitives Problems: Require standardization of application interactions dedicated to the IoT IoT app store are missing Recommendations: IoT platforms must provide SDKs and APIs that maximize the re-usability of the services provided by their platform Julien Mineraud Overview of IoT Platforms / 37

20 Ecosystem formation Gap analysis of IoT plaforms Current status: Platforms provide useful building blocks, storage and run-time environment for application developers Expectations: Platform easily expandable by the developers Cross-platform sharing of applications and services Local composition of services Gaps: Low platform expandability Limited monetizing possibilities Limited support for cross-platform integration Problems: Silos of platform-specific solutions User s using multiple platforms experience difficulties for integration Recommendations: A broker is needed to ease cross-platform integration Models to contextually define IoT applications to simplify their discovery by the end-users Julien Mineraud Overview of IoT Platforms / 37

21 IoT marketplace Gap analysis of IoT plaforms Current status: Limited applications sharing Limited (usage-based) charging of the end users of these applications Expectations: Dedicated IoT data catalogs, IoT app store and IoT device store Advertise, deliver and charge for the use of applications and data Validate applications against policies Gaps: Application, data and device catalogs dedicated to the IoT are generally missing The billing of the end-users of the data is generally missing Problems: An ecosystem of independent application developers, device manufacturers, and end-users all supporting the platform is needed for the demand for marketplace to appear and sustain Recommendations: The marketplace functionality shall be provided by future IoT platforms Julien Mineraud Overview of IoT Platforms / 37

22 Gap analysis of IoT plaforms Summary of the gap analysis Results: IoT platforms are mainly cloud-based Focus is on accessibility (sensors available from anywhere at anytime) Numerous protocols supported The interoperability between IoT platforms is extremely limited Motivation is to break the verticalization of IoT solutions: More standardization of IoT protocols and models (e.g., onem2m) MQTT, CoAP, LWM2M, IPSO etc. More interoperability in applications and plugins Not all IoT data can be processed in the cloud Problem of transportation, storage, security & privacy We need to enable the local processing of IoT (edge analytics) Create web of IoT platforms: Innovation for services New business models (IoT Marketplace) Julien Mineraud Overview of IoT Platforms / 37

23 Gap analysis of IoT plaforms Recommendations for the development of future platforms Leaning on standardized communication protocols to interface heterogeneous devices Adding the provisions for handling and processing data locally Adding uniform data models, data catalogs, and the edge analytics capabilities Offering streamlined APIs Introducing cross-platform brokers and financial incentives for ecosystem players Developing dedicated IoT marketplace(s) Julien Mineraud Overview of IoT Platforms / 37

24 Gap analysis of IoT plaforms A vision with the IoT Marketplace Julien Mineraud Overview of IoT Platforms / 37

25 Gap analysis of IoT plaforms Immediate benefits of the IoT Marketplace Data processing & sharing: theabilitytorequestnumerousexternal data streams to enrich local content. It would also enable users to publish some of their streams to third-parties Developer support: thepossibilityforapplicationdevelopersto publish their products and reach a wide range of customers Ecosystem formation: theincreasingawarenessaboutnew innovations and possibility of creating new business models Market & billing: theabilitytomarket/searchfordataand applications and sell/purchase the rights to use them Julien Mineraud Overview of IoT Platforms / 37

26 Useful IoT platforms Outline 1 Introduction 2 Evaluation of the IoT platform landscape 3 Gap analysis of IoT plaforms 4 Useful IoT platforms Open-source platforms Another interesting platform 5 IoT Hub & IoT Market Julien Mineraud Overview of IoT Platforms / 37

27 Useful IoT platforms Open-source platforms Afewopen-sourceplatformsI node-red ( Node-RED is an open-source Node.js tool that aims to simplify the connection between IoT devices and web services. It incorporates the concept of flow for IoT devices and data that allows complex interactions between objects and services. The flow can be published on the NodeRED website for sharing. The thing system ( The thing system is a software using Node.js that enables discovery of smart things in the home environment. The project is open-source and licensed under the M.I.T license. The software does not provide storage functionalities and must be coupled with a PaaS to enable storage outside the home area. The software intends only to provide access remotely to smart devices of smart homes. Julien Mineraud Overview of IoT Platforms / 37

28 Useful IoT platforms Open-source platforms Afewopen-sourceplatformsII Ericsson IoT-Framework ( The Ericsson IoT-Framework is a PaaS that accumulates sensor data from IP networks and focuses on the analytics and the mashing up of the data. The strength of this platform is the publish/subscribe mechanism, and querying of data streams, both from local and external data sources) to perform analytical tasks ThingSpeak ( ThingSpeak is decentralized, open-source and provides a server that may be used to store and retrieve IoT data. It allows opening of the channels (data flows) but do not provide extensive configuration of the data flows. The platform also provides visualization tools and enables the creation of widgets Julien Mineraud Overview of IoT Platforms / 37

29 Useful IoT platforms Another interesting platform Another interesting platform IFTTT ( ( if this then that ) is a SaaS offering, allowing a rapid composition of services called recipes by applying simple if-then rules to external service building blocks, such as s, Facebook event that either play the role of atrigger(if)oranaction(then,do).therecipescanbepersonalorshared at the discrepancy of the user Julien Mineraud Overview of IoT Platforms / 37

30 IoT Hub & IoT Market Outline 1 Introduction 2 Evaluation of the IoT platform landscape 3 Gap analysis of IoT plaforms 4 Useful IoT platforms 5 IoT Hub & IoT Market Julien Mineraud Overview of IoT Platforms / 37

31 IoT Hub & IoT Market IoT Hub The IoT Hub has been developed in the DIGILE Internet of Things program & is envisioned as an extension of existing IoT platforms to support interoperability IoT Hub API REST API to access IoT elements Devices, data, services & application Flexibility in access control Solves the data ownership issues Script engine High modularity All components of IoT Hub are scripted Used for plugins, services and applications Will be integrable to all IoT platforms with IoT Hub extension Julien Mineraud Overview of IoT Platforms / 37

32 IoT Hub & IoT Market Exposing your data with the IoT Hub One can connect devices to an IoT Hub, but to use them, one must create an external representation of them, known as the IoT feed Feeds Feeds are of three kinds Atomic feeds Connected to IoT devices or external services via scripted plugins Atomic feeds only represent one basic type of data (abstracted using standard models such as IPSO) Composed feeds More traditional, on the fly definition of multiple data types (e.g., temperature + humidity + location) Executable feeds Exposes IoT Hub VM (Javascript) to third parties for distributed computing Julien Mineraud Overview of IoT Platforms / 37

33 IoT Hub & IoT Market Accessibility of IoT feeds The IoT Hub includes the support of roles that defines: How the feeds can be accessed (read, write, execute) Which functions from the feeds are available Summary, count, etc.. The IoT Hub also includes the support of users Assigned to a set of profile Hub owner is having the admin role (everything is authorized) Authentication is done via the OAuth 2.0 protocol Login with username, password! token Use token to make requests on the hub API Julien Mineraud Overview of IoT Platforms / 37

34 IoT Hub & IoT Market IoT Market We envision to have particular type of IoT hubs used as marketplaces and catalogs to create web of IoT platforms Just another IoT Hub Data is not from sensors but from other hubs Services & applications Link to IoT Hub applications and plugins Interconnect IoT Hub instances Exposing the IoT Hub services and the roles required to access them Decentralized approach Need for dedicated marketplaces (e.g., transport, urban data, etc.) Julien Mineraud Overview of IoT Platforms / 37

35 IoT Hub & IoT Market IoT Hub & IoT Market Two-level architecture Markets and hubs IoT Market (a.k.a. meta-hub) A front-end to applications and services provided by the hubs AIoTsearchengineusingaglobalIoT index IoT Hub Core of the data management (data privacy, fusion & sharing) Connectivity to external devices IoT Hub API REST API for accessing hubs and IoT Market Julien Mineraud Overview of IoT Platforms / 37

36 IoT Hub & IoT Market Edge analytics with the IoT Hub Challenges One of the biggest challenges introduced by the Internet of Things is to analyze the tremendous volume of data that is being produced by IoT devices in real-time. The current solution to analyze data is to all of it to the cloud. However, this may raise a number of privacy and security issues. IoT Hub analytics The IoT Hub platform allows the platform owner to authorize a special kind of service to external services: IoT Hub script engine. This enables IoT Hub platforms to cooperate and organize as micro-clouds to perform analytics very close to the edge of the network (i.e., where the data has been produced) Julien Mineraud Overview of IoT Platforms / 37

37 Additional materials Additional materials IoT platforms survey analysis.html IoT Hub Video: Source code (Apache 2.0): kahvihub (Java and Android) node-iothub (Node.js) Julien Mineraud Overview of IoT Platforms / 37