You are here›Blog›Bringing Stream Analytics to M2M-Based SmartLife Scenarios
Bringing Stream Analytics to M2M-Based SmartLife Scenarios
Ericsson Spain hosts a major R&D Center with development responsibility over Data Management and Traffic Management products, an Ericsson Research branch and a local Business Lab (an organization sponsored by the market region aiming at bringing innovations to the market). This set-up creates obvious synergies. On one hand, the Ericsson Business Lab receives market requirements and insights, especially from areas such as utilities, transport and government, while Ericsson Research provides cutting-edge technologies that can be tested and deployed in different market scenarios.
Ericsson is a member of the consortium created to carry out the BUTLER project through the R&D Center in Madrid. BUTLER* (uBiquitous, secUre inTernet-of-things with Location and contEx-awaReness) is an EU FP7 Integrated Project that focuses on the Internet of Things (IoT).
The main focus of BUTLER is IoT, enabling the development of secure and smart life applications relying on context and location-aware, pervasive information systems. To that end, BUTLER will:
- Define different use cases in different scenarios (Home, Office, Transportation, Health, and Shopping). However, BUTLER’s proposal is different because it focuses on specific technologies of context-awareness that work pervasively across multiple situations, instead of focusing on specific scenarios.
- Integrate/developing a new flexible smartDevice-centric network architecture where platforms (devices) function according to three well-defined categories: smartObject (sensors, actuators, gateways), smartMobile (user’s personal device) and smartServers (providers of contents and services), interconnected over IPv6.
- Build a series of field trials, which progressively integrate and enhance state-of-the-art technologies to showcase BUTLER’s secure, pervasive and context-aware vision of IoT.
The Ericsson Contribution
Ericsson’s contribution relies on the strengths of the Ericsson R&D Center in Madrid, both technically and from a business perspective. The technical contribution takes advantage of the expertise built on Stream Analytics. Stream Analytics is a set of technologies focused on real-time analysis such as Complex Event Processing (CEP), sometimes referred to as Data Stream Management, or Stream Mining. CEP focuses on analyzing streams of data (an ordered sequence of instances that can be read only once or a small number of times using limited computing and storage capabilities) from multiple sources and deriving a conclusion from them on the fly without requiring the storage of data. Stream Mining is a subset of Data Mining and, in the same way as CEP, extracts knowledge from streams of data as they pass, instead of being applied off-line.
However, technical expertise must be complemented with business value. It is at this point that the business innovation skills of the Business Lab join the team. Through its involvement with the market region operations, the Business Lab is bringing requirements and business analysis from the utilities and government areas. This guarantees that our involvement in BUTLER relates to actual business opportunities.
Ericsson is starting its contribution to BUTLER by building on top of the proofs of concept that we have developed using Complex Event Processing technologies. Our target is to use them as a basis for implementing the use cases defined in the project.
The proofs of concept are the following:
- Distributed Generation with Renewable Energy Sources (for the SmartHome scenario) and
- Real-Time Mobile Advertisement (SmartShopping).
Distributed Generation with Renewable Energy Sources
This proof of concept addresses a scenario where renewable energy sources are deployed at the users’ homes. Such sources are usually photovoltaic solar panels, but they could be also microwind turbines. The use of renewable energy addresses environmental concerns. Furthermore, when the sources are located at the user's premises, energy consumption is much more efficient. However, renewable energy collection is not as predictable as traditional energy generation, so that if this uncertainty could be minimized, the cost of energy management would be decreased and therefore the energy efficiency will increase over time.
We propose the monitoring of renewable energy sources at home premises using on-line and off-line tools to build a model for the prediction of the energy production, taking into account different environmental variables (e.g. weather variables). By doing this, a utility can have more precise estimates about the amount of renewable energy that the customer is generating in the near future at home, and of their needs in terms of energy consumption. On one hand, the utility can plan better the generation of energy using non-renewable sources. On the other, home appliances tasks can then be tentatively scheduled according to periods where renewable energy production is high.
Real-time data analysis technologies (CEP) are used to forecast both the in-site energy generation and consumption to make the energy balance and use more efficient.
This proof of concept will be contributed to the BUTLER SmartHome scenario.
Real-Time Mobile Advertisement
This proof of concept addresses a scenario where information related to the location of a user and to the web content s/he is accessing is used to make real-time decisions on information or advertisement the user is offered.
The Real-Time Mobile Advertisement monitors data streams captured from an operator’s network and from external sources to send targeted ads to subscribers of a telco network.
Potential consumers can receive ads on their mobile devices as they approach physical shops surrounded by “ads zones”. These ads are targeted to the specific subscriber by matching information collected from the operator’s network. The ad zone may be, for example, one or several coverage areas of cells in the cellular network of the operator. The ads are sent in real-time as a person passes (walking, driving, etc.) near (from meters to kilometers) a physical shop.
This scenario enables new use cases in a telecom network through CEP technology. It allows advertisers to send targeted ads according to user interests, profile and context (mainly location) to consumers that have previously opted-in. While information on interests and profile is mainly static, the context is continuously extracted from the analysis of the user location and the web pages and applications the user is accessing or using, by analyzing the streams sent to the CEP system by the proxy handling web traffic at the operator network. Ad matching is performed close to or in the access network, allowing the reuse of location infrastructure and lower latencies.
This proof of concept will be contributed to the BUTLER SmartShopping scenario.
More information can be found here.
Besides the aforementioned proofs of concept, Ericsson is contributing the Device Data eXchange (DDX) concept to the BUTLER project. DDX is a proposal for the monetization of real-time sensor data in smart cities environments. This cloud-based service will be provided to device/data owners on one hand, and to application developers on the other. Device/data owners can get further revenue from the M2M infrastructure they have already deployed by letting the DDX operator handle the engagement with data consumers and charging. They can also benefit from positive externalities such as economic stimulation, service innovation or end-user engagement (see the positive externalities that comprehensive horizontal ICT infrastructure provides in smart cities in “Information Marketplaces: the New Economics of Cities”). Developers are offered a one-stop shop where they can access a large amount of real-time data, provided through simple APIs and without the need to worry about how data consumption is charged (see value chain in slide 16 of “Building Cities for the Networked Society”).
One of the main challenges of the research and innovation projects is to link them to actual business opportunities. The Ericsson R&D Center in Madrid has been working in stream processing techniques for more than two years and has developed a deep understanding of the real-time analysis technologies. However, technical mastery is not enough if it is not possible to provide value to customers. Real-time analysis technologies are suitable in different areas: black-out and fraud detection in utilities, energy demand-response scenarios, real-time advertisement and recommendation, detection of abnormal or unexpected human behavior (emergencies, unusual weather conditions, demonstrations…), real-time exposure of sensor data within smart city scenarios for the creation of new services, preemptive maintenance of industrial equipment. All of these are areas with strong market demands which can be supported by stream analytics.
~ Miguel-Angel Monjas & Manuel Lorenzo, Ericsson Research, Spain
* BUTLER was kicked off on October 1st, 2011 with a planned duration of 36 months. The BUTLER consortium comprises 17 partners from 8 different countries: academic institutions and public research institutes (K.U. Leuven, iHomeLab, University of Oulu, etc.) and companies (Telecom Italia, ST Microelectronics, Gemalto, Swisscom, Ericsson, etc.). The total budget of the project is around €14.9M.