Patent Application
20240013892
The present invention relates to a system and method for monitoring the habits of a user. In more detail, the system and method of the invention are also intended to generate a set of food and housing instructions in an optimized form for the user.
From the post-war period onwards, energy requirements have grown in proportion to the increase in population density.
In the light of the drastic climate change taking place, a number of international agreements aim to reduce per capita energy use to sustainable averages without changing people's quality of life. One such agreement is that developed by the Swiss Federal Institute of Technology in Zurich and called the “2000 Watt Society”, which aims to reduce the energy needs of each individual to a maximum of 2000 W and 1 tonne of energy per person of CO2 by 2100. The aim is to count not only the personal energy used—or the dwelling energy consumed—but also the grey energy, i.e. the amount needed to produce, transport and dispose of a product/service or material.
In this context, the design of future dwellings also plays a key role in the 2000 W society project. Many intelligent housing solutions are known in which energy is produced using exclusively renewable sources combined with the use of improved home automation in order to increase the overall energy efficiency of the dwelling unit. However, although these solutions improve in many respects and aim to increase self-sufficiency, they are in fact mostly confined to automated energy management of the dwelling without actually caring about the health and long-term energy balance of the occupants.
In the light of the objectives set out above and the attempts at energy improvement developed by the prior art, the Applicant has realized that in this historical period of constant climatic deterioration, it is no longer sufficient to optimize only the envelope and/or the energy efficiency of the dwelling. In a broader vision of an even more sustainable energy balance, it is necessary to focus more on the inhabitant and his or her behavior. The habits of human beings, their way of living, their health and mobility are therefore extremely decisive in calculating the overall energy balance, their well-being in general, in order to achieve a real climate improvement.
The Applicant has therefore decided to develop a system that will make it possible to monitor not only the energy values of the buildings but also the behavior of the users who live in them as part of an extremely high-performance energy balance project that will, at the same time, have a profound effect on the way future homes are built. In particular, the monitoring of habits will make it possible, in a predefined time frame, to instruct the user on new optimized ways of living, moving and eating, acting within the framework of food self-sufficiency related to the building.
By virtue of the above, one of the objects of the present invention relates to a system for monitoring the habits of a user which permits sending optimized instructions from both a dietary and a habitation point of view in order to measure and count in a more precise and efficient way the real energy need of an individual within the target of 2000 W society.
The Applicant also thought that the provision of the optimized instructions should include a combined analysis of the dwelling's energy production data, the users' nutritional information as well as the constant analysis of their physiological parameters corresponding to both movement data and the liquid and solid evacuations produced by the user over time.
By virtue of the foregoing considerations, this and other objects of the present invention are achieved by the system for monitoring the habits of a user according to claim 1.
Another object of the present invention is to devise a method for monitoring the habits of a user and to generate a plurality of optimized dietary and/or habitation instructions comprising the use of one or more neural networks according to claim 7.
Further characteristics and advantages of the system and method according to the present invention will result from the description below of preferred embodiments thereof, given by way of an indicative yet non-limiting example, with reference to the attached figures, wherein:
With reference to the attached illustrations, U is globally referred to as a user meant to inhabit a dwelling D and to be monitored by the system of the present invention. As will be seen in the remainder of the present disclosure, some embodiments of the present invention have the objective of instructing and helping the user U (or possibly several users U of the same dwelling D) in different areas, such as, e.g.:
As will be seen below, the instructions and suggestions as well as the household automations that will be provided by the system of the invention will be used for the purpose of calculating the overall energy balance of the user during his or her daily life, outside and inside his or her dwelling, in order to maintain as much as possible a lifestyle that allows an average annual energy demand of about 2000 W within the target of 2000 W Society.
Globally, the energy requirements and CO2 emissions of the dwelling D as well as of the user U are set to meet the limit target values, preferably by the date 2100, of approximately 2000 W per capita. The dwelling D should have the highest possible degree of self-sufficiency compatible with the need to limit the energy used during its construction (so-called grey energy).
With reference to the example in
With reference to the example in
By way of example, the second physiological parameters P2 comprise parameters related to the user's vital values such as: protein, glucose, urine ph, liters of urine, etc.
Advantageously, the system has processing means M configured to receive the parameters P1, P2, process them and generate at least one food instruction IA and/or dwelling instruction ID optimized for the user.
In the remainder of the present description and in the subsequent claims, “food instruction” means one or more pieces of nutritional information suggested on the type of food the user should use to feed himself or herself according to the detected physiological parameters P1, P2 and other parameters which will be better specified in the remainder of the present description. Similarly, “dwelling instruction” means one or more pieces of information for the automatic and non-automatic coordination of the user points of the dwelling as well as units for the production of renewable energy such as heat, electricity and/or water.
Currently, the average energy demand (primary energy) per person in Europe is basically as follows:
As will be seen later in this disclosure, the parameters P1, P2 are used by the processing means M to calculate an average annual energy requirement E of the user U equal to or less than 2000 W (e.g. equal to 5.5 W per day or 2 kWh per hour or 48 kWh per day). To this end, one of the aims of the invention envisages that the dwelling D be completely self-sufficient in terms of food; it is therefore possible to have one or more plant organisms 3 or cultivable crops inside or outside the dwelling D. Such plant organisms 3 may comprise, e.g., fruit plants, able to provide not only sugary fruit but also complex carbohydrates, fats and proteins, berries with a triple attitude (fruit, biomass, nitrogen fixation), vegetables, other annual and biennial plants, carbohydrate and protein crops (potatoes, cereals, legumes) and other crops important for the self-supporting food and health of the user.
The design of the layout of the plant organisms 3 associated with the dwelling D is set up with the aim of minimizing processing and ensuring adequate health from the point of view of the phytosanitary profile of the crops as well as ensuring the best possible use of space without overloading the nutrient cycle in the soil with excessive drawdowns. In essence, feeding is compared and considered as an energy requirement of the building on a par with any other energy produced. The Applicant has ascertained that by means of a personal plant-based diet based on seasonal and home-grown products, it is possible to obtain a primary energy value of approx. 325 MJ/m2 with expected greenhouse gas emissions of 12 kg/m2.
Preferably, for each plant organism 3 or for each area containing one or more plant organisms, suitable detecting devices 4 can be provided, configured to acquire nutritional values P3 regarding, e.g., the growth and/or the state of maturity of the foods produced by them, their nutritional value, or even other readings such as soil pH, soil fertility (quantity of nitrogen, phosphorus, potassium, magnesium, calcium, sulfur), quantity of organic substance, etc. The value P3 of the foods will then be sent to the processing means M, according to known transmission techniques, in order to analyze them and coordinate them with the physiological parameters P1, P2 received from the same processing means M and consequently update the food instructions IA and/or dwelling instructions ID in order to optimize them.
According to one version, the detecting devices 4 may comprise dedicated sensors such as, e.g., cameras, biosensors, etc. in order to acquire values P3. Conveniently, cameras may be used, either inside (
According to one embodiment, the system comprises at least one unit for the production of renewable energy 5. Preferably, such unit 5 may comprise devices for the production of energy by means of wind, geothermal, hydroelectric, marine, and/or solar systems. The unit for the production of renewable energy 5 may therefore comprise, e.g., solar/photovoltaic panels, water turbines, windmills, etc.
In the example shown in
The electrical exchange values P4 may also comprise information about stored energy, energy produced by the panels, energy consumed, grey energy, etc. Additionally, the values P4 may comprise values of energy absorbed by the household appliances such as, e.g., absorption (in kWh), noise (dB), liters of water consumed, etc.
According to one embodiment, the user points 7 of the system may comprise at least one vehicle 8 of the smart type for the transport of the user U such as, e.g., a car, a bicycle, a motorbike, etc. The vehicle 8 comprises at least one control unit 9 for the calculation of the mobility values P5 such as, e.g., consumed electrical values, emissions, distances traveled, noise, etc., and for the communication of these values P5 to the processing means M using known transmission techniques, such as, e.g., transmission of the wireless type.
Preferably, the vehicle 8 is an electric type vehicle provided with its own internal rechargeable battery. The battery of the vehicle 8 is rechargeable via suitable charging station 10 associated with the dwelling D.
Conveniently, the mobility values P5 are sent to the processing means M and used by the latter to coordinate them with the physiological parameters P1, P2, the nutritional values P3 and the electrical exchange values P4 in order to update the food instructions IA and/or dwelling instructions ID to be provided to the user U in an optimized manner.
It should be observed that the food instructions IA and/or dwelling instructions ID to be given to the user are processed by the processing means M taking into account the average energy demand of approximately 2000 W of user U, as well as of the dwelling D, calculated over a period of one year. It may happen in fact that in some periods of the year, for example during the winter seasons, the average energy demand exceeds 2000 W. In this context, the processing means M will be configured to optimize consumption by balancing it over the year, e.g. by instructing the user U to consume less during summer periods.
Conveniently, the food instructions IA may comprise indications as to the type of food the user should consume on a daily basis, e.g. through particular diets. In particular, the instructions IA generated by the processing means M take into account both the food consumed by the user U by means of the productions from the plant organisms 3 of the dwelling D, and the food consumed by the user U not produced by such organisms 3. If, in fact, the user U dines out (e.g. in a restaurant), the user will send information about the food consumed to the processing means M by means, e.g., of a dedicated app installed on his or her smartphone.
According to a preferred embodiment, the system of the invention has a smart virtual assistant 11 in signal communication with the processing means M in order to interact with the user U. The assistant 11, by means of the processing means M, is able to receive information from all the devices installed therein such as, e.g., the detecting means and devices 1, 2, 4, the different units for the production of renewable energy 5, the user points 7, the vehicles 8, computers, notebooks, smartphones, etc. and to process the parameters P1-P5 thereof for the purpose of generating the food instructions IA and/or the dwelling instructions ID.
In one version, the processing means M are mounted on board the virtual assistant 11.
In a further version, the system may have a plurality of audio interfaces of the combined microphone/speaker type (not shown) in each room of the dwelling D, or also installed at predefined points outside, in signal communication with the personal assistant 11 in order to be able to acquire voice commands from the user U and emit the food instructions IA and/or the dwelling instructions ID in audible format.
In yet another version, the food instructions IA and/or the dwelling instructions ID can also be given through different interfaces, e.g. via the smartphone of the user U, in text format via dedicated apps, etc.
As shown in the example of
In detail, the training of the neural network N is carried out through one or more preliminary phases of training used for the setting up of a training dataset through, first of all, an initial screening of the user U and of the dwelling D. Such screening allows acquiring a plurality of distinguishing characteristics df(U), df(D) that characterize the user U and the dwelling D. The training of the neural network N, aimed at learning the distinguishing characteristics df(U), df(D) as well as the information and properties associated with the user U and with the dwelling D, is carried out according to predefined machine learning algorithms. The training of the neural network N can be e.g. carried out through determinative algorithms (e.g. CNN, YOLO, SSD, etc.) able to learn not only the decision function but also the different representations of the distinguishing characteristics df(U), df(D) at different information levels.
In detail, in one of the preliminary training phases, there is a phase of providing a plurality of initial information about, e.g., the general health status of the user U and a plurality of properties/attributes concerning the devices and crops associated with the dwelling.
By way of example, the screening of the user U may comprise one or more of the following activities acquired at the beginning of the training phase:
Subsequently, some of the above screening activities can be repeated at predefined monthly or annual intervals.
Similarly, the screening of the dwelling D allows defining a plurality of distinguishing characteristics df(D) taking into account at least one of the following characteristics:
Preferably, it must be specified that the characteristics df(D), df(U) related to the above listed characteristics are and may be selected and designed in a way to represent as much as possible qualitative technicalities always in step with the state of the art actually present on the market.
According to one version, the evolution and the state of the aspect related to food productions will be constantly monitored during the training phase and documented in order to produce adequate training of the neural network N on the productions to reach the prefixed objectives at the level of coverage of the food needs of the user U.
It has been estimated that a first phase of training can include a period of time equal to about three years during which the neural network N and, at the same time, the users U acquire knowledge in the field of horticulture, field-crops, processing, storage, etc. During this first phase, the soil, fertility in the broadest sense, and the various biological and microbiological balances of the natural system around the dwelling site will also be in the midst of a stabilization phase, so that the procurement of the distinguishing characteristics df(U), df(D) relating to this period will focus on monitoring the evolution and stabilization of the system rather than on actually achieving the maximum expected goals.
A second training phase which goes, e.g., from the third to the sixth year will make it possible to improve overall knowledge of the dwelling D as well as the agronomic knowledge of the soil so that the neural network N and consequently the user U achieve sufficiently consolidated learning. In this time frame, shrub and tree crops generally start to give satisfactory yields. In this second phase, the monitoring carried out by the processing means M guided by the neural network N will be of essential importance as it will allow generating food instructions IA and/or dwelling instructions ID for the user U which are increasingly more optimized in order to make possible a degree of energy and food self-sufficiency, thus consolidating the coordination between the values and the parameters P1-P5 acquired and constantly monitored.
A third phase of training, for instance from the sixth year on, will allow reaching a fine tuning level so that the food instructions IA and/or dwelling instructions ID generated by the processing means M, in association with the optimized decisional functions of the neural network N, allow calibrating the energy consumptions and the food productions to maintain an annual average energy requirement of about 2000 W within the target of 2000 W Society without affecting the wellbeing of the user U.
The above-described learning will allow the neural network N to assign, for each value and parameter P1-P5 and/or for each determined distinguishing characteristic df(U), df(D) acquired, a predefined accuracy threshold on the basis of the distribution of the analyses conducted during the preliminary training phases in order to minimize possible false positives/negatives while maintaining, at the same time, a high capacity to provide an immediate response during the subsequent phases of execution of the present invention, as will be seen later in the present disclosure.
From what has been said above, it is clear that the training is intended to set a training dataset containing all the information and the representative associations of the identified distinguishing characteristics df(U), df(D), as well as of the single parameters P1-P5, which are constantly saved in a database connected to the processing means M to train the neural network N.
It must be specified that already by the end of the procurement phase described above, it will be possible to interrogate the neural network N to command the execution of the processing means M and therefore generate one or more food instructions IA and/or dwelling instructions ID for the user U. Furthermore, during the entire training phase, the neural network N is both able to learn and to execute the instructions IA, ID in continuous mode. In point of fact, the neural network N improves the learning by adapting the execution and improving it as the training proceeds. In other words, the neural network N is substantially always in continuous training.
In summary, the system of the present invention provides for a plurality of training phases intended for the training of the neural network N which make it possible to automatically generate one or more food instructions IA and/or dwelling instructions ID optimized for the user U within the target of the 2000 W Society.
With reference to the examples shown in the attached illustrations, shown below are some of the possible phases of execution and generation of the food instructions IA and/or dwelling instructions ID according to the present invention.
In order to better understand the multitude of functions that the system of the invention can perform, the different operating conditions must be defined of the user U and/or of the dwelling D.
In relation to the user U, at least the following operating conditions can be defined, also combinable with each other:
With reference instead to the dwelling D, at least the following operating conditions can be defined, also combinable with each other:
By way of example, a typical day of an occupant/user U of the dwelling D could start by receiving a series of instructions suggested to him or her by the virtual assistant 11 comprising, e.g., dwelling instructions IA (morning wakeup) and food instructions ID (the food to be taken from the pantry for breakfast). The wakeup call will be set automatically by the neural network N on the basis of the habits of the user U acquired over time during the training, according to the statistical average rest requirement and age, etc. On the other hand, the instructions IA concerning breakfast will also be suggested automatically by the virtual assistant on the basis of the training of the neural network N, depending on the food requirements of the user U, of the products available from the crops of the dwelling D, of the average annual energy requirement E, of the parameters P1-P5, etc.
Similarly, the mode of use of the vehicles 8, as well as of the user points 7 will be suggested by receiving one or more dwelling instructions ID depending on the energy produced by the dwelling D, the average annual energy requirement E, the parameters P1-P5, etc.
If the user should have lunch or dinner outside the home, the virtual assistant 11, for example through his or her smartphone, will send a plurality of food instructions IA on the type of food to be consumed always on the basis of the eating habits of the user U which the neural network N already knows. It follows that the latter will adapt its decisional functions by taking into account the annual average energy requirement E equal to or below 2000 W.
In virtue of the above, it has been ascertained that the calculation of the average energy requirement E can depend on different factors, for instance on:
In view of this, the neural network N is configured to generate one or a plurality of algorithms governing the conditions/phenomena under consideration on the basis of a predefined function constituted by the processing of all the information listed above.
As has been verified by the present description, it has been ascertained that the invention described herein achieves the intended objects and in particular the fact is underlined that by means of the system described herein it is possible to generate, with extreme accuracy, a plurality of food instructions IA and dwelling instructions ID for the user in a speedy automatic and optimized way within the scope of the 2000 W Society.
Furthermore, thanks to the innovative execution phase, it is possible to use the same instructions IA, ID to train the neural network N continuously over time. The effectiveness of the decisional functions of the neural network N thus gradually improves over time and speeds up the generation and effectiveness of the instructions IA, ID.
Finally, it should be noted that if the neural network N is not sure to provide an accurate result, it is always possible to question experts (doctors, nutritionists, etc.) so that the information of the neural network N can always be updated even with monitoring not acquired by the system.
Several tests have been carried out in order to simulate the dwelling conditions of a user and the supply of the optimized food instructions IA and/or dwelling instructions ID to the user to maintain an energy consumption of less than 2000 W per capita. The tests were carried out using the SimaPro software that allows collecting, analyzing and monitoring the environmental performance of products and services according to the ISO 14040-14044 standards. For the input of the values of the parameters P1-P5, df(U) and df(D), reference was made to some known data (e.g. nutritional values, energy consumption, physiological needs, etc.) while others were calculated using various databases, especially regarding environmental impacts such as:
By way of example, the main databases used were Ecoinvent, USA Input Output Database, LCA Food DK, USLCI, ELCD, Industry data.
The tests made it possible to establish different user profiles and different instructions IA, ID which allow maintaining an average annual energy requirement E of about 2000 W per user. A planet/user sustainability ratio of one planet per 2000 W was also defined, as shown in the example in
A score of “A” indicates excellent behavior that would contribute to the sustainability of the planet, a score of “G” would indicate bad behavior whereby consumption would contribute to the exploitation, over time, of as many as five Earths. Thanks to the present invention, finally, it is possible to obtain a real climatic improvement by considering all the consumptions related to a user without neglecting any of them; the combination of the characteristics of the monitoring system as well as its execution are potentially endless and obviously a technician in the field, in order to satisfy contingent and specific needs, may make numerous modifications and variants, all of which are contained within the scope of protection of the invention, as defined by the following claims.
| Number | Date | Country | Kind |
|---|---|---|---|
| 102020000020197 | Aug 2020 | IT | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/IB2021/057127 | 8/4/2021 | WO |
