Periodic Reporting for period 1 - dSense (dSense – Self adapting, cost efficient method for detecting context of a mobile device and a mobile device with a context detection module)
Berichtszeitraum: 2016-05-01 bis 2016-10-31
Zusammenfassung vom Kontext und den Gesamtzielen des Projekts
The problem
There are ways of specifying a mobile device context, based on analysis of signals from different sensors (microphones, accelerometers, light sensors, magnetic field sensors, compasses, cameras, etc.), using global satellite positioning system (GPS), signals of base stations of cellular telephony and calculating average speed of the device based on the change of thus determined position in time. Another alternative, used on an industrial scale, is to use vehicle-mounted short-range radio transmitter. There are also other solutions providing method of detecting a mobile device context, but they all deliver low accuracy of results or they require direct engagement of the user or they use of significant amounts of energy.
Therefore, currently in the industry are carried out intensive developments of effective and efficient methods for determining the context based on signals from various types of sensors. This leads often to the situation of different types of sensors in parallel, delivering one type of information- e.g. that the user is in move. Our method is directed to meet compromise between accuracy and time precision of mobile applications’ results and minimizing energy consumption of batteries operating in mobile devices. Additionally, our algorithm will be a self-adapting system, thanks to which less energy consuming sensors will be ready to successfully detect a context, appropriately to individual user’s behavior and his customs.
Why is it important for society?
At the end of 2013 ca. 60% of mobile applications’ users said that they use their apps daily. Smartphones apps are mostly used for: search, portals & socials; entertainment; communication; productivity & tools; commerce & shopping . We have identified four areas of needs according to understanding of the user:
• First group consists of final users of applications and mobile devices, who are strongly interested in solutions which give them max. useful results, in an appropriate time (accuracy and speed), using minimum of energy, what directly regards to device’s usability and costs of its operating.
• Second group consists of application developers/investors, who will want more precisely get to the potential customer. That will lead to optimization of the balance between marketing, delivery, operating etc. costs, and revenues from properly targeted customers and classification of their needs.
• Third group consists of mobile devices producers which mount the module of context detection module in their devices, who will want be more attractive on the market (for end-user, as well as application developers), what will bring better economic results for the company.
• Fourth group consists of MEMS Combo Sensors producers, who will want be more attractive on the market, what will bring better economic results for the company.
In addition, the project will contribute to the Europe 2030 Strategy in its main points, i.e. intelligent (increase of high-technology role in Europe’s economy) as well as sustainable development (moving towards reduce greenhouse gas emissions by at least 40% compared to 1990 levels and a 27% increase in energy efficiency).
Overall objectives
The final outcome of the project is to launch in existing mobile devices a software system and/or module for a context detection to ensure environmentally and costly efficient required service’s results for users.
The main innovative feature of technology is new algorithm of collaboration among sensors, mounted in mobile devices, to maximize the effectiveness of compromise between precision of the results and energy consuming in mobile applications being in everyday use. Today’s mobile devices are equipped with a large number of different types of sensors, which allow, inter alia, automatic activation and deactivation of individual functions or to change configuration of mobile devices, depending on the context.
Current solutions, as they are not self-learning, may return incorrect results, leading to excessive energy consumption or poor user experience. Our self-learning/-adapting solution will allow better experience for the user, and also will not waste the energy, as it happens currently.
There are ways of specifying a mobile device context, based on analysis of signals from different sensors (microphones, accelerometers, light sensors, magnetic field sensors, compasses, cameras, etc.), using global satellite positioning system (GPS), signals of base stations of cellular telephony and calculating average speed of the device based on the change of thus determined position in time. Another alternative, used on an industrial scale, is to use vehicle-mounted short-range radio transmitter. There are also other solutions providing method of detecting a mobile device context, but they all deliver low accuracy of results or they require direct engagement of the user or they use of significant amounts of energy.
Therefore, currently in the industry are carried out intensive developments of effective and efficient methods for determining the context based on signals from various types of sensors. This leads often to the situation of different types of sensors in parallel, delivering one type of information- e.g. that the user is in move. Our method is directed to meet compromise between accuracy and time precision of mobile applications’ results and minimizing energy consumption of batteries operating in mobile devices. Additionally, our algorithm will be a self-adapting system, thanks to which less energy consuming sensors will be ready to successfully detect a context, appropriately to individual user’s behavior and his customs.
Why is it important for society?
At the end of 2013 ca. 60% of mobile applications’ users said that they use their apps daily. Smartphones apps are mostly used for: search, portals & socials; entertainment; communication; productivity & tools; commerce & shopping . We have identified four areas of needs according to understanding of the user:
• First group consists of final users of applications and mobile devices, who are strongly interested in solutions which give them max. useful results, in an appropriate time (accuracy and speed), using minimum of energy, what directly regards to device’s usability and costs of its operating.
• Second group consists of application developers/investors, who will want more precisely get to the potential customer. That will lead to optimization of the balance between marketing, delivery, operating etc. costs, and revenues from properly targeted customers and classification of their needs.
• Third group consists of mobile devices producers which mount the module of context detection module in their devices, who will want be more attractive on the market (for end-user, as well as application developers), what will bring better economic results for the company.
• Fourth group consists of MEMS Combo Sensors producers, who will want be more attractive on the market, what will bring better economic results for the company.
In addition, the project will contribute to the Europe 2030 Strategy in its main points, i.e. intelligent (increase of high-technology role in Europe’s economy) as well as sustainable development (moving towards reduce greenhouse gas emissions by at least 40% compared to 1990 levels and a 27% increase in energy efficiency).
Overall objectives
The final outcome of the project is to launch in existing mobile devices a software system and/or module for a context detection to ensure environmentally and costly efficient required service’s results for users.
The main innovative feature of technology is new algorithm of collaboration among sensors, mounted in mobile devices, to maximize the effectiveness of compromise between precision of the results and energy consuming in mobile applications being in everyday use. Today’s mobile devices are equipped with a large number of different types of sensors, which allow, inter alia, automatic activation and deactivation of individual functions or to change configuration of mobile devices, depending on the context.
Current solutions, as they are not self-learning, may return incorrect results, leading to excessive energy consumption or poor user experience. Our self-learning/-adapting solution will allow better experience for the user, and also will not waste the energy, as it happens currently.
Arbeit, die ab Beginn des Projekts bis zum Ende des durch den Bericht erfassten Berichtszeitraums geleistet wurde, und die wichtigsten bis dahin erzielten Ergebnisse
The main objective of the study was to assess the technical feasibility and economic viability of the implementation of dSense on the global market. The work completed under the project was included in:
1. the Feasibility Study – includes an overview of the work completed and the analysis of the technical part of the implementation,
2. the Business Plan – includes the financial part, the strategic part and the market part of SME I project.
Additionally, when working on the above-specified documents, the following were created:
• The Technological Audit Report and the plan of development stages, that contain the report of conclusions from the analysis of the technical aspects and possibilities of dSense commercialization in the form of a public API
• The Financial model (dynamic Excel model).
Above mentioned documents allowed the Company to reach the specific objectives that were defined for the "SME I" project, i.e.:
• Identification of potential technical challenges which should be addressed before or during industrial research stage;
• Identification of crucial resources needed to success of next stage of product development;
• Assessment of market entry opportunities and suitable business models
• Examining financial viability;
• Potential partners identification.
Above listed specific objectives were being achieved by activities done within the SME I project.
Main results:
Our innovation is very efficient and very required solution on the market, what is proved by many attempts of finding context detection solutions made by global and regional markets players, which create trends of development of mobile applications. Currently, these applications support or have potential to support every sector of industrial, scientific, social, entrepreneurial life. The economy tends to bring faster, more accurate answer to identified needs of its various markets' players. It is especially well seen in trends of Industry 4.0 advertising solutions, social media, etc. Prompt and apt support of every-day operations is required both by the demand and supply site of the market.
As the market of sensors is developing dynamically, there is always a high need of improving their collaboration with applications which they support. In the future we would like also to explore and develop possible methods of exploiting and expand our solution to other devices, what could open new markets for Binartech.
Conclusions of the action are as follows:
• The target market is absorptive and growing;
• dSense responds to the identified needs of the distinguished target groups;
• Competitive advantages of dSense has been identified;
• In the range of intellectual property management, portfolio management methodology was implemented;
• Work plan and budget for next steps of development is realistic and feasible;
• Binartech has sufficient human and financial resources needed to market implementation of dSense. Infrastructure gaps has been identified and included in the financial analysis;
• All analyzed economic ratios indicated the viability of the implementation;
Relying on the works performed, the materials collected and the conclusions which were drawn on that basis and included the relevant risk, the Company has decided to continue the dSense project.
1. the Feasibility Study – includes an overview of the work completed and the analysis of the technical part of the implementation,
2. the Business Plan – includes the financial part, the strategic part and the market part of SME I project.
Additionally, when working on the above-specified documents, the following were created:
• The Technological Audit Report and the plan of development stages, that contain the report of conclusions from the analysis of the technical aspects and possibilities of dSense commercialization in the form of a public API
• The Financial model (dynamic Excel model).
Above mentioned documents allowed the Company to reach the specific objectives that were defined for the "SME I" project, i.e.:
• Identification of potential technical challenges which should be addressed before or during industrial research stage;
• Identification of crucial resources needed to success of next stage of product development;
• Assessment of market entry opportunities and suitable business models
• Examining financial viability;
• Potential partners identification.
Above listed specific objectives were being achieved by activities done within the SME I project.
Main results:
Our innovation is very efficient and very required solution on the market, what is proved by many attempts of finding context detection solutions made by global and regional markets players, which create trends of development of mobile applications. Currently, these applications support or have potential to support every sector of industrial, scientific, social, entrepreneurial life. The economy tends to bring faster, more accurate answer to identified needs of its various markets' players. It is especially well seen in trends of Industry 4.0 advertising solutions, social media, etc. Prompt and apt support of every-day operations is required both by the demand and supply site of the market.
As the market of sensors is developing dynamically, there is always a high need of improving their collaboration with applications which they support. In the future we would like also to explore and develop possible methods of exploiting and expand our solution to other devices, what could open new markets for Binartech.
Conclusions of the action are as follows:
• The target market is absorptive and growing;
• dSense responds to the identified needs of the distinguished target groups;
• Competitive advantages of dSense has been identified;
• In the range of intellectual property management, portfolio management methodology was implemented;
• Work plan and budget for next steps of development is realistic and feasible;
• Binartech has sufficient human and financial resources needed to market implementation of dSense. Infrastructure gaps has been identified and included in the financial analysis;
• All analyzed economic ratios indicated the viability of the implementation;
Relying on the works performed, the materials collected and the conclusions which were drawn on that basis and included the relevant risk, the Company has decided to continue the dSense project.
Fortschritte, die über den aktuellen Stand der Technik hinausgehen und voraussichtliche potenzielle Auswirkungen (einschließlich der bis dato erzielten sozioökonomischen Auswirkungen und weiter gefassten gesellschaftlichen Auswirkungen des Projekts)
Our innovation is based on the sensors’ work, mounted in the all mobile devices. Identified problems of energy consuming sensors are the main problem, that we would like to tackle. Various sensors used in mobile devices have different level of their technical advancement according to their role.
For example, some sensors can identify the movement of the mobile device (understood as a movement of the user), but not all of them can precisely say if the person in fact is moving from point A to B (eg. driving a car), what could be important for applications connected with localization identification, etc.
On the other hand, sensors used in applications to identify e.g. localization, directions of movement and connected to that other circumstances and user’s requirements, use a lot of energy, although in many cases results they return are not useful for the user. What is worse, many times, various sensors which could identify the context of the user in the same or similar way, work in parallel, according to applications that they support.
The above described approaches are costs’ and energy insufficient. We want to develop the model of sensors’ division according to their energy consumption, and support their communication between less and more energy consuming sensors. The problem of reducing the energy consumption, is typically solved by less frequent switching the sensors on and less frequent sampling. This leads however to extension of time after which a change of device context is detected.
However, thanks to our invention, less advanced sensors could as first try to identify the context, and, after getting a positive impulse, contact more advanced sensors just to confirm or deny accuracy of the result. After the positively detected context, the more advanced sensors will be switched on, to support the user with their services, giving at the same time a feedback to this “lower” sensors, if their results of context definition was successful or not.
That will open an opportunity to teach less advanced sensors to “read” the context appropriately, and eliminate false identification of the context next time in similar situation. It will directly influence three key aspects of usage of mobile application: their accuracy, and energy consumption, influencing positively costs' decrease, as well as self-adapting mobile application to specific needs of the user.
The novelty of the project relates to a method for detecting a context of a mobile device equipped with sensors and a context detection module, and its self-adapting feature, in which the sensors are assigned to at least two groups, each of which comprises at least one sensor, and each group is allocated a group of classifier adapted to detect, in a form of a classification result, currently identified, by means of given classifier, context device based on the indications of the sensors belonging to the given group.
With the use of a context detection module, the groups of sensors are ordered hierarchically, the device context is detected by reading a classification result indicated by the classifier or the currently active group, wherein in case of detection of an identified context in the active group, there is switched on power supply of the sensors and there is activated classification in a group with higher level and there is read the context indicated by said group’s classifier, wherein based on the results of the classification indicated by the higher groups classifiers there is made an adaptation of the configuration of lower groups’ classifiers.
Impact
Unique benefits the new product will provide to prospective customers:
Increased energy efficiency of mobile devices.
Thanks to this optimization it will also contribute to ensure better energy efficiency of batteries used in these devices. It will be possible thanks to our innovation focused on the method for detecting a context of a mobile device equipped with various sensors, as well as on a context detection module. The project outcome will supply a ready solution of how to use existing and planned to use sensors mounted in the mobile devices. Our method focuses on division of sensors into various groups, according to their energy efficiency and accuracy of outcomes, and building the system of their interrelationships, to optimize the applications’ results and total energy consumption.
Faster and apt fit of mobile services to user’s needs.
Today’s mobile devices are equipped with a large number of different types of sensors, which allow, inter alia, automatic activation and deactivation of individual functions or to change configuration of mobile devices, depending on the context. A special case of the context is a situation when a user of a mobile device drives a vehicle, such as a car. It is inadvisable in such circumstances, and in many countries prohibited by law to use such devices’ functions so as make and receive voice calls and send and receive text messages. On the other hand it may be advantageous to activate other functions, such as navigation or download form an external database of information on known hazards on the roads and to inform the user of approaching them, with a message of a tone, voice, visual or any combination thereof.
Currently, mobile application are designed without any auto-control of accuracy and the quality of offered services. Many times their activity is not welcomed by the user, or even may disturb the user during important, hazard or discrete situation. It is not acceptable, but it also should not require from the user to deactivate all applications before this specific situation. The mobile device is expected to read properly the context of the user, and propose him/her services appropriate to the situation. For example, turn on the GPS sensor and location information, when it is required by the user or, when his/her movement in a vehicle is detected. Current solutions, as they are not self-learning, may return incorrect results, leading to excessive energy consumption or poor user experience. Our self-learning/-adapting solution will allow better experience for the user, and also will not waste the energy, as it happens currently.
For example, some sensors can identify the movement of the mobile device (understood as a movement of the user), but not all of them can precisely say if the person in fact is moving from point A to B (eg. driving a car), what could be important for applications connected with localization identification, etc.
On the other hand, sensors used in applications to identify e.g. localization, directions of movement and connected to that other circumstances and user’s requirements, use a lot of energy, although in many cases results they return are not useful for the user. What is worse, many times, various sensors which could identify the context of the user in the same or similar way, work in parallel, according to applications that they support.
The above described approaches are costs’ and energy insufficient. We want to develop the model of sensors’ division according to their energy consumption, and support their communication between less and more energy consuming sensors. The problem of reducing the energy consumption, is typically solved by less frequent switching the sensors on and less frequent sampling. This leads however to extension of time after which a change of device context is detected.
However, thanks to our invention, less advanced sensors could as first try to identify the context, and, after getting a positive impulse, contact more advanced sensors just to confirm or deny accuracy of the result. After the positively detected context, the more advanced sensors will be switched on, to support the user with their services, giving at the same time a feedback to this “lower” sensors, if their results of context definition was successful or not.
That will open an opportunity to teach less advanced sensors to “read” the context appropriately, and eliminate false identification of the context next time in similar situation. It will directly influence three key aspects of usage of mobile application: their accuracy, and energy consumption, influencing positively costs' decrease, as well as self-adapting mobile application to specific needs of the user.
The novelty of the project relates to a method for detecting a context of a mobile device equipped with sensors and a context detection module, and its self-adapting feature, in which the sensors are assigned to at least two groups, each of which comprises at least one sensor, and each group is allocated a group of classifier adapted to detect, in a form of a classification result, currently identified, by means of given classifier, context device based on the indications of the sensors belonging to the given group.
With the use of a context detection module, the groups of sensors are ordered hierarchically, the device context is detected by reading a classification result indicated by the classifier or the currently active group, wherein in case of detection of an identified context in the active group, there is switched on power supply of the sensors and there is activated classification in a group with higher level and there is read the context indicated by said group’s classifier, wherein based on the results of the classification indicated by the higher groups classifiers there is made an adaptation of the configuration of lower groups’ classifiers.
Impact
Unique benefits the new product will provide to prospective customers:
Increased energy efficiency of mobile devices.
Thanks to this optimization it will also contribute to ensure better energy efficiency of batteries used in these devices. It will be possible thanks to our innovation focused on the method for detecting a context of a mobile device equipped with various sensors, as well as on a context detection module. The project outcome will supply a ready solution of how to use existing and planned to use sensors mounted in the mobile devices. Our method focuses on division of sensors into various groups, according to their energy efficiency and accuracy of outcomes, and building the system of their interrelationships, to optimize the applications’ results and total energy consumption.
Faster and apt fit of mobile services to user’s needs.
Today’s mobile devices are equipped with a large number of different types of sensors, which allow, inter alia, automatic activation and deactivation of individual functions or to change configuration of mobile devices, depending on the context. A special case of the context is a situation when a user of a mobile device drives a vehicle, such as a car. It is inadvisable in such circumstances, and in many countries prohibited by law to use such devices’ functions so as make and receive voice calls and send and receive text messages. On the other hand it may be advantageous to activate other functions, such as navigation or download form an external database of information on known hazards on the roads and to inform the user of approaching them, with a message of a tone, voice, visual or any combination thereof.
Currently, mobile application are designed without any auto-control of accuracy and the quality of offered services. Many times their activity is not welcomed by the user, or even may disturb the user during important, hazard or discrete situation. It is not acceptable, but it also should not require from the user to deactivate all applications before this specific situation. The mobile device is expected to read properly the context of the user, and propose him/her services appropriate to the situation. For example, turn on the GPS sensor and location information, when it is required by the user or, when his/her movement in a vehicle is detected. Current solutions, as they are not self-learning, may return incorrect results, leading to excessive energy consumption or poor user experience. Our self-learning/-adapting solution will allow better experience for the user, and also will not waste the energy, as it happens currently.