Building Management Through AI

Building Management Through AI

Shrey Jain

School of Architecture, IPS Academy IPSA, Indore, India

Arjun Sharma

School of Architecture, IPS Academy IPSA, Indore, India

Abstract – Imagine a building that behaves like Jarvis in Iron Man, like knowing when to turn on the heat in the house, before you feel cold, or turning on the lights as soon as you enter a room, or making life easier for disabled people. This is what this study is about, its about using artificial intelligence AI and sensors in buildings.

Today, buildings consume large amounts of energy, accounting for as high as 40 percent of global energy consumption. On the other hand, over 1.3 billion disabled people face difficulties living within buildings that lack sufficient accommodations for them. Existing building systems work based on standard programming that doesn’t offer any flexibility or learning capability. The purpose of this paper is to show the benefits of using modern technology based on artificial intelligence and sensors in buildings. Sensors function similarly to a nervous system for the building, detecting any unusual movement and sending information about it to the central processor.

The potential use of these technologies is illustrated through examples like adaptive facade technology that changes according to weather condition, predictive heating/cooling, lighting that adjusts according to the task being performed, and security systems that identify you based on face recognition. Most importantly, the use of artificial intelligence systems in helping people with disabilities through voice control, detecting falling incidents, mobility support systems, and customized systems for specific individuals are discussed here.

It may take some time to build the full-fledged system but there are technologies available now that can be used to accomplish this mission. The article provides a very pragmatic approach to developing such systems and highlights the challenges that would be faced including privacy of personal data, coordination between various subsystems, accessibility issues, and user acceptance.

Keywords- Artificial intelligence, Smart Buildings, Sensors, Building Management System etc.

1. INTRODUCTION

   In the Marvel Universe, Tony Starks intelligent assistant named Jarvis performs much more functions other than answering queries. The assistant controls the entire building, predicts the need for any function even before it occurs, tracks everything at all times, and provides rapid emergency responses faster than a person. Although Jarvis was created by an imaginary character from the movie universe, the reality is that such an assistant can now be achieved due to advances made in the fields of artificial intelligence, smart sensors, and building automation technology. In this research work, we look at how such an integration of AI and sensor technologies can help in designing and implementing intelligent buildings, which will become more energy-efficient and accessible to disabled persons as well.

   Buildings have always had an additional importance beyond being mere physical entities. The environment in which humans interact with one another and carry out everyday activities determines many aspects of health, efficiency, well-being, and comfort. However, despite their critical importance, the intelligence of today’s buildings is relatively low. The schedule of operations is fixed; changes in conditions are ignored; and the behavior of buildings does not depend on specific individuals using it. Heating starts at a certain time irrespective of the occupancy level. Lights burn in empty offices; and security allows access to everyone or blocks access entirely. These shortcomings are not only inconvenient but are also quite costly. Buildings consume 40% of all energy produced globally. This means that buildings are the biggest consumers of energy in the world.

   It should be noted that this problem affects not only energy consumption. The number of people who have various disabilities around the globe exceeds 1.3 billion individuals (World Health Organization, 2011). For them, movement across a common building turn into an impossible mission, as these people are not able to open the difficult doors, adjust lightening conditions, operate complex environmental controls, and receive notifications from emergency systems due to poor vision. That means that the constructed environment does not care about the diversity of people.

   The conventional BMS were initially created in order to manage particular processes and use energy efficiently. However, these devices lack ability to learn something. For instance, a system may be set in order to turn off the lights after the last person has left a room, but it is not possible for it to recognize the fact that a worker requires more time to get out of a building at the end of his/her working day or a person may need more time until the automatic doors are closed.

   AI creates a totally new situation altogether. It is able to detect patterns of behavior of the occupants and respond accordingly. Visionary abilities allow it to detect who the occupant of the room is and what activities he or she is engaged in. Language processing allows it to communicate with the environment in natural language, meaning that no physical interfaces would be required for this communication. Predictive analytics makes it possible to anticipate needs before the occupant realizes them.

   The introduction of AI will bring about a totally different approach. The algorithms will be able to understand and interpret the behavioral pattern of occupants and respond appropriately. Computer vision will detect whose presence is in the room and determine what activity he is carrying out. The language processor will enable communication with the environment through language with no need for any interface whatsoever. Predictive analysis will predict needs before they are known by the occupant.

   This paper explores the way in which all these technologies can be incorporated into a building management system after the lines of Jarvis. Innumerable possible applications are considered, such as kinetic facade, predictive heating ventilation air conditioning systems, artificial intelligence – based lighting, and accessibility through artificial intelligence. This paper will take up some of these actual barriers which would have to be overcome, including data security, technical feasibility, economic efficiency, and making sure that technology helps where it should help.

   Not that this paper intends to explore mere theoretical potentialities of contemporary technology. The aim of this paper is to lay out a viable scheme for intelligent building management system, one that is both efficient and humane. Our life revolves around buildings; it is high time they started taking lessons from us.

2. METHODOLOGY

   1. Research Design

      1. Research Design

         In terms of the methodology used in this study, it relies on the interaction of two different research methodologies employed in a complementary manner. As opposed to a laboratory experiment that would focus on building a model that demonstrates some particular property of the studied system, this research involves not only the construction of a framework, but also analysis of prior findings of other researchers and an examination of currently available technologies. At the same time, these elements are combined in order to develop a framework for an intelligent buildin system that resembles the system used by Jarvis. The choice to employ this methodology stems from the fact that the sphere of research is relatively young, and the big picture provides more solid ground than an experiment.

         It should be noted that the following four steps will be involved in the research process:

         Step 1 Analysis of existing studies and technologies Step 2 Elaboration of the system requirements

         Step 3 Development of the framework and architecture

         Step 4 Assessment of the proposed framework’s capabilities and weaknesses

      2. Read Research and Evaluate Technology

         The next step involved the reading of various research articles, technical reports and case studies relating to the following topics: AI/ML in building environments

         Smart sensors and Internet of Things Kinetic and adaptive building facades Predictive heating and cooling control

         Assistive technologies for people with disabilities Privacy and security in smart buildings

         Building accessibility standards and universal design

         The sources included articles published in reputable journals such as Energy and Buildings, Applied Energy, IEEE Transactions on Industrial Informatics, and Pervasive and Mobile Computing. There were also technical reports by various international bodies including the WHO, International Energy Agency and ASHRAE. A total of 35 sources were evaluated to determine viable technologies, current gaps in technology, and emerging trends applicable to this project.

         The technologies identified were classified according to maturity level as currently available, developmental or aspirational, in order to have a realistic and accurate assessment of available technologies.

      3. Understanding What the System Should Accomplish

         In order to create the right solution, it was necessary to establish an understanding of what the system should accomplish. This was evaluated considering two aspects, i.e., what the building requires and what its users require.

         Requirements for building:

         Constant monitoring of all spaces throughout the entire time period Energy saving while not affecting people’s comfort

         Compatibility with existing building infrastructure Fault detection and recovery capabilities Compatibility with both small and large buildings People’s requirements:

         General users – comfort, convenience, security, awareness regarding energy usage

         Disabled people – assistance in movement, alternative communication methods for vision or hearing-impaired users, voice control, hand gestures recognition, fall detection, emergency notifications, personalized environment configuration that will keep user’s individual settings

         As always, I have to admit that there was a certain bias in this part of my research disability requirements were taken into account based on the literature I managed to find rather than interviewing disabled people directly.

      4. Designing the Framework and System Architecture

         The key achievement of this study is that it provides the design of the proposed integrated AI building management system. This system has been developed as a modular framework in such a way that each module is independent but interconnected with each other through the AI brain of the system.

         The proposed system is structured as five interrelated layers:

         Layer 1 The Sensor Network

         This layer is the base of the whole system and consists of the physical devices:

         Environmental sensors monitoring temperature, humidity, CO2, air quality, and illumination Occupancy sensors identifying occupancy status in each space of the building

         Motion sensors detecting motion, abnormal activity (falls), unusual behavior (stillness for too long)

         Biometric sensors identification via fingerprint, face recognition, voice authentication (to set personalized parameters)

         Smart cameras ensuring the safety, counting number of people inside a room, assisting users with disabilities through computer vision

         Structural sensors identifying fires, leaks, vibrations, state of exterior surfaces of the building facade All of these data from sensors are then transferred to the next layer using special communication methods used by smart devices, including MQTT and Zigbee, as these protocols provide energy efficiency and speed.

         Layer 2 Local Data Processing

         At the local processing stage, all sensor information will be processed before being uploaded to the cloud at the next layer using special processors located within the building. The purpose of this step is to process any data in real time when required; for example, detecting the person’s fall or alarming about a fire. Moreover, performing local processing also prevents private data from being leaked as only aggregated data is transferred to the cloud for further training purposes.

         Layer 3 The AI Brain

         This is the most crucial layer and one that is most influenced by Jarvis. It comprises several specialized AI modules, including:

         The Prediction Module, which uses past patterns to make predictions about future circumstances in order to make preemptive adjustments

         The Learning Module, which creates personalized profiles of all permanent residents by collecting information about the preferences for things such as the ideal temperature and lighting

         The Voice Control Module, allowing anyone to command the building through speech, thus being particularly useful to those unable to operate switches and buttons physically

         The Computer Vision Module, which uses video feeds to recognize faces, falling accidents, number of people inside the building, and hand signals

         The Fault Detection Module, which keeps an eye on all systems and gives early notifications when something seems off to avoid malfunctions

         The Energy Management Module, always looking for a balance between the comfort of inhabitants and energy efficiency in real time

         Layer 4 Building Systems Control

         In this level, the AI brain is responsible for implementing decisions taken by making the physical building operate based on those decisions. Thus, all systems that include heating and cooling systems, lighting, door access control, kinetic façades, elevators, emergencies, and assisting devices will be controlled using the building management system. Moreover, the building systems feed the brain about their state continuously.

         Layer 5 How Humans Will Communicate and Control the Intelligence of the Building

         This is a level that includes all the means by which humans will interact and control the buildings intelligence: Mobile application to check and adjust settings remotely

         Voice commands accessible anywhere in the building

         Management dashboard for use by building managers and maintenance teams Accessibility control interface that will be specially designed for disabled individuals Emergency alert system using audible, visual, and vibrating signals

      5. Testing and Evaluation of the Proposed Framework

         The following four issues were considered when testing the effectiveness of the framework:

         Can it be implemented? Is the required technology already available?

         Does it conserve energy? What is the potential for achieving energy savings? Does it help disabled persons? Does it really solve their problems?

         Is it secure and confidential? Does it conform to existing data protection requirements?

         These four issues were answered through comparison of the proposed framework with actual smart buildings. Several such smart buildings, located in Singapore, Germany, the Netherlands, and the USA, were analyzed via case study, and whenever possible the actual results of their work were used to assess the capabilities of the proposed framework.

         All the difficulties encountered in the course of this test are described openly in th following part of the paper.

      6. Ethics

         Considering that the system involves collecting sensitive information such as face identification data, movements, and personal health data, including incidents like falling, ethics were incorporated as one of the core components of the solution from the start of the development process, not as an afterthought. The ethical framework used five main criteria:

         Collect only what you need never collect unnecessary data

         Always obtain consent residents will always be made aware of what is being collected and why Local storage preferred sensitive data remains within the building wherever possible

         Respect the right of self-determination each occupant can review, amend, and delete their personal profile anytime Inclusivity for all each function of AI technology was thoroughly examined and ensured to work for disabled people

3. KINETIC FACADE MANAGEMENT

   1. Definition of Kinetic Facade

      Think of the façade of the building as the moving skin of an organism. Unlike regular buildings whose wall surfaces and windows do not respond at all to the changes in the surrounding environment, a kinetic façade is quite different from that as it is a movable or flexible façade comprising of panels or shading devices. In essence, it could be compared to the scales of a fish or the petals of a flower.

      Kinetic facades, as stand-alone structures, can be regarded as great engineering feats. However, combining them with AI that is powered by weather sensors, solar radiation sensors, and occupancy sensors turns this into something extraordinary. In essence, a building becomes capable of choosing appropriate attire for different weather conditions, much like a person chooses what clothes to wear depending on the weather outside.

   2. Control of the Kinetic Facade Using Artificial Intelligence

      If a kinetic facade were to work without the use of artificial intelligence, the control mechanisms for such systems would include the activation of panels to shade based on light intensity when it reached a certain threshold. Such an approach would work, although it would not be very efficient. With the help of artificial intelligence, however, the system would become extremely efficient due to the ability to collect information about several aspects at once.

      This includes light intensity from solar sensors, weather information such as wind and temperatures from weather sensors, room temperature and humidity from room sensors, and whether a room is occupied or not. The use of all that information will allow making precise adjustments and preparing the building for upcoming conditions. For example, with the help of weather predictions, the AI will begin cooling the building during the morning hours to make it easier to cope with the high temperatures in the afternoon.

   3. Real World Applications

      Al Bahr Towers, Abu Dhabi This building design is based on 1,049 umbrella-shaped shades, which are modeled after Islamic screen designs. All 1,049 elements move independently depending on the position of the sun, helping reduce solar gains by about 50%, and thus significantly reduce energy used for air conditioning.

      Kiefer Technic Showroom, Austria The building includes 112 independent aluminum panels that can be moved through a computerized control system, according to their preference in terms of shade and lighting conditions.

   4. Energy Saving and Cost Issues

   There is much room for saving energy with an AI-controlled kinetic facade. The studies reveal that a smart kinetic facade system allows reducing cooling energy usage by 30-60% in tropical climate zones (Loonen et al., 2013). In addition to energy saving, users receive benefits of increased natural light exposure, elevated mood and productivity, and zero glare issues.

   Kinetic facades are costly as they may increase construction expenses for a building’s envelope by up to 15-40%. However, considering the entire building lifetime of 30-50 years, the savings on energy and improved work productivity usually compensate for the increased construction costs within 8-15 years.

   For cheaper constructions with less energy-saving goals, there is always a possibility of implementing AI-controlled motorized blinds/louvres which provide a good balance between effectiveness and cost-efficiency.

4. AI FOR PERSONS WITH DISABILITIES

   1. Importance of the Topic

      There are more than 1.3 billion people in the world who are living with disabilities (World Health Organization, 2011). Many individuals find things like getting into the building, adjusting the temperature, and making calls in an emergency situation difficult. The reason for this is that the buildings are designed keeping in view the needs of an average person who does not have any kind of physical limitation. AI can transform this scenario by providing such buildings which can adjust according to the individual’s needs without asking or struggling for anything.

   2. Voice Control and Smart Interfaces

      Voice control is another area where artificial intelligence is of great assistance to the physically challenged individuals. For example, accessing the light switch or adjusting the thermostat may require physical effort that some physically challenged individuals cannot make because of their conditions. Using an artificial intelligent interface that is integrated in the building, all one requires is to talk to the building, and the building performs tasks such as turning on lights, opening and closing doors, raising the temperature or pulling down blinds.

   3. Detection of Fall Incidents and Response

      Falls represent one of the biggest threats that elderly persons, as well as those who suffer from issues with balance, are exposed to in their homes. The system uses artificial intelligence-based sensors and camera systems to detect movement around the property in real time. Once there is a sudden change in the movement, which indicates a fall, the system alerts the appropriate emergency contact immediately, regardless of whether the person can request assistance independently.

   4. Assistance for Visually Impaired and Hearing-Impaired People

      In the case of visually impaired people, the buildings AI system would assist with providing audio-guidance assistance, increasing lighting intensity wherever the visually impaired individual travels, and detecting obstacles by using sensors that issue an alert to the individual about the obstacle using either sound or vibrational cues via their smart device.

      In the case of hearing-impaired individuals, all auditory alerts will be replaced by visual and vibration cues. When there are fire alarms, the building will light up in certain areas. Similarly, when a doorbell rings, the message will show on a screen or vibrate the smart device of the occupant.

   5. Personalized Learning for Each Occupant

   Undoubtedly, one of the most amazing aspects of such AI is its capability to learn about the requirements of each and every occupant of the building. It knows what light intensity a certain occupant likes, how long it should keep the doors open and when to start operating the elevator in order for it to be at the destination when the user arrives there. The system builds up personalized data for every occupant and then uses it whenever the user enters the building. For the handicapped population, it would mean having a really comfortable environment without the necessity to adjust it manually or seek help from someone else.

5. ENERGY MANAGEMENT AND SUSTAINABILITY

   1. Energy ineficiency in buildings

      Energy Issue in buildings are amongst the major consumers of energy in the world consuming almost 40 percent of the total energy consumed in the world (Pérez-Lombard et al., 2008). Most of this energy goes to waste because people forget to switch off the lights when there are no people in the room, turn off heating and air conditioners, and keep running without any need to be adjusted to fit the current condition of the room. Building management systems use static methods that are incapable of being intelligent enough about what is going on in the building.

   2. Predictive Heating and Cooling

      The energy consumed for heating and cooling purposes is normally about half the total energy consumed in the building. AI does this through identifying the occupancy patterns of individuals and predicting the time and areas where heating or cooling is necessary. Instead of heating the entire building beginning at 8 am because the majority of the occupants arrive at this time, AI analyzes the occupancy data at the moment, the weather forecast, and past trends to determine which parts of the building need to be heated and the exact time. It has been found that AI-based HVAC optimization may lead to up to 30% decrease in energy consumption without compromising the level of comfort (Javed et al., 2017).

   3. Smart Lighting Control

      Lighting is yet another significant contributor to wasted energy in buildings. AI-controlled lighting uses occupancy detectors, day light detectors, and machine learning algorithms to make sure that lights are only turned on where and when they are really required. In the presence of daylight, artificial lighting automatically adjusts and reduces its brightness. When the room is vacant, lights turn off almost instantly. All of these factors result in a lot of energy savings over the whole year for the entire building.

   4. Integrating Renewable Energy

      In addition, AI has a key part to play in the management of renewable sources of energy. For instance, AI observes the production of energy from solar power on the roof in real time and estimates the total amount of energy produced through solar panels for the entire day through the weather forecast. After that, high energy consumption activities, such as dishwashers or even pre-cooling rooms, are automatically performed when the energy produced from solar panels is at its maximum.

   5. A Broader Perspective

   All of these components working together within the ambit of AI management produce quite amazing effects. Studies conducted on actual smart buildings have indicated that they save anywhere between 20% and 40% energy compared to traditionally managed buildings (Molina-Solana et al., 2017). In addition to being economically efficient, it makes the building a part of the solution to climate change rather than being an aggravating factor through its carbon footprint.

6. CASE STUDIES

   Real world examples of smart buildings:

7. FUTURE SCOPE AND RECOMMENDATIONS

   1. Where Is This Technology Going?

      Building	Location	Key AI Feature
      The Edge	Amsterdam	World’s smartest office building
      Al Bahr Towers	Abu Dhabi	Kinetic facade system
      Bullitt Center	Seattle	Net zero energy AI management
      Marina One	Singapore	AI environmental control
      The Crystal	London	Fully integrated smart BMS

      While the current state of AI building management technologies is truly remarkable, it is but a prelude to what is yet to come. The ideal case, where a self-sufficient Jarvis-style system would anticipate all the needs of its master, provide perfect management of all processes, and even feel almost as another intelligence in the house, seems still out of reach. But the rapid evolution in AI technologies, sensors, and computing makes it much closer to reality than one might expect at first glance.

      A number of revolutionary technologies are emerging already. For instance, brain-computer interfaces, developed by such companies as Neuralink, might eventually let totally paralyzed individuals fully control all the building processes with mere thoughts.

      AI technologies are also progressing towards the point where they can identify medical emergencies prior to their full manifestation. In the future, buildings may have capabilities that enable them to sense the beginnings of a heart attack or stroke in a person by means of biometric technology and contact medical professionals in advance.

      On a broader scale, there is potential for the interlinking of smart buildings with each other as well as with other smart city elements, such as electricity management and emergency response systems.

   2. Recommendations

   To achieve this vision, cooperation between researchers, developers, architects, and policymakers will be required. It is crucial to invest in economical sensors so that these technologies can be applied in more places than luxurious buildings. Accessibility guidelines mandating that AI-enabled assistance be included in any newly constructed public building should be developed by governments. The most important thing is the inclusion of disabled people in the design process.

8. DISCUSSION

   1. How Far Are We from the Jarvis Model?

      This paper demonstrates that even though a completely autonomous Jarvis-like building automation system has yet to be created, we are much further along toward realizing this goal than most people think. Each of the component technologies necessary for its creation artificial intelligence, sensors, voice command systems, machine vision, predictive analytics have been successfully implemented and tested in a variety of buildings worldwide. What remains is simply tying them all together into one cohesive system.

   2. Who Benefits Most?

      From what can be deduced from the results of this study, the most beneficiaries of AI-powered management of buildings are not the rich office goers in luxurious smart buildings, but rather the people suffering from disabilities, old occupants, and those living in places characterized by severe climatic conditions. It is important to note this fact since the discussion of smart buildings revolves much more around luxury than social inclusion and human needs.

   3. Ethical and Social Implications

   By far, the biggest issue to consider regarding the management of the construction of intelligent buildings is ethical rather than technological. A building that tracks its occupants via sensors and cameras, observes behavioral patterns, and records biometric data is definitely one that keeps tremendous amounts of personal information. In the absence of adequate safeguards and cybersecurity measures, it is easy to turn the very thing that benefits people into a method of surveillance and oppression. This should be taken into account when moving ahead with further development.

9. CONCLUSION

1. Summary of Major Findings

   The objective of this study was to find out whether the idea of an AI building management system inspired by Jarvis would be possible to implement in the real world. Having looked at all of the issues raised in this paper, it seems clear that such implementation would indeed become possible not yet perhaps, but certainly with the technology that we currently have. From smart buildings whose exeriors adapt to changes in weather conditions, to buildings that employ AI to provide autonomy to physically disabled residents, there is clear evidence that intelligent buildings will soon be our reality.

2. Responding to the Research Aim

   The purpose of this research was met when the ability of AI technologies working in conjunction with distributed sensors was proven as a means to improve energy efficiency and comfort while also offering an essential contribution towards the lives of individuals living with disabilities. All critical components of this suggested architecture can be found in existing technology.

3. Conclusion

The buildings are what make humans experience life every single day. They dictate the level of comfort, amount of energy consumed by the Earth, and independence of individuals with disabilities. Creating buildings that can think, learn, and really care for those who stay within them is not only a question of technology, but also a matter of ethics. We have all that we need its time to start building.

REFERENCES

1. Buckman, A. H., Mayfield, M., & Beck, S. B. (2014). What is a smart building? Smart and Sustainable Built Environment, 3(2), 92-109.

2. Javed, A., Larijani, H., Ahmadinia, A., & Gibson, D. (2017). Smart random neural network controller for HVAC using cloud computing technology. IEEE Transactions on Industrial Informatics, 13(1), 351-360.

3. Loonen, R. C., Trka, M., Cöstola, D., & Hensen, J. L. (2013). Climate adaptive building shells: State-of-the-art and future challenges. Renewable and Sustainable Energy Reviews, 25, 483-493.

4. Molina-Solana, M., Ros, M., Ruiz, M. D., Gómez-Romero, J., & Martín-Bautista, M. J. (2017). Data science for building energy management: A review. Renewable and Sustainable Energy Reviews, 70, 598-609.

5. Pérez-Lombard, L., Ortiz, J., & Pout, C. (2008). A review on buildings energy consumption information. Energy and Buildings, 40(3), 394-398.

6. World Health Organization. (2011). World report on disability. World Health Organization.