Home Automation System

Sustainability CodeCamp - Spring 2016

Prof. Dr. Olaf Drögehorn

​

Group 6: Giang Tran, Nhi Vo, Aigerim Zhalgasbekova, Chandara Chea, Henrique Sarmento, Victor Araujo

Our happy team

Outline

Part 1: Our Project

1. Motivation
2. Goals
3. Solution
4. Implementation
5. Results
6. Limitations and future works
7. Demo

Part 2: THREAD Protocol

​

Heating & Ventilation Control System

Part 1:

1. Motivation

• People all have bad habits in terms of using electricity
• HVAC accounts for the largest amount of energy use in homes
• Automatically control the HVAC system based on the user's occupancy can save up to 30% of total energy consumption
• Cultivate good behaviour in using energy

2. Goals

• Save energy used by HVAC system in home environment
• Maintain the comfort environment for users automatically
• Foster good behaviour by using data collected from home automation system

3. Our solution

Automatically control the radiators and ventilation based on users' occupancy status in the room to ensure the comfort level in the room and save energy

�

3. Our solution

• Using machine learning to improve the efficiency of our solution
• FHEM is used to feed a datastore accessible through an API for machine learning or creating other services on top of it.
• A neural network was trained to attempt to learn a fictional schedule.�

Why machine learning?

A Lawrence Berkeley NL, UC Davis and UC Berkeley study in 2010 found that:

“many occupants used the PT as an on-off switch and most demonstrated little knowledge of how to operate it. The on-line survey found that 89% of the respondents rarely or never used the PT to set a weekday or weekend program. The photographic survey (in low income homes) found that only 30% of the PTs were actually programmed.”

U.S Department of Energy (2015):

“You can easily save energy in the winter by setting the thermostat to 68°F while you're awake and setting it lower while you're asleep or away from home. By turning your thermostat back 10° to 15° for 8 hours, you can save 5% to 15% a year on your heating bill.”

​

​

3. Our solution

FHEM

• Raspberry PI

HomeMatic

• Adapter LAN
• Motion sensor
• Door/window sensor
• Temperature and Humidity sensor
• Radiator Thermostat

​

3. Our solution

4. Implementation

• The system will detect the occupation status inside the room.
• If user are in the room, the temperature will be maintained in comfort levels
• If user leave their room for work but other users are still in their room, maintain the comfort temperature
• If there is no one inside the room during working time, decrease the temperature and if it reaches a certain level, shut it down. (scheduling not implemented)
• During the night, if there is no movement, decrease the temperature (scheduling not implemented)

4. Implementation

• If user feels too cold or hot, they can adjust the temperature to their desired levels. When they leave their house they can turn down the heating (not implemented).

​

• The system will learn it and apply these knowledge to control the heating system by applying a more precise desired temperature. (We can train a machine learning algorithm with FHEM data. Next step would be to control the devices back)

​

5. Result

• Assumption: 1 student apartment, 2 radiators
• Without the home automation system:

1500(W) * 2(heaters) * 24(h) * 30(days) = 2,160 kWh

• With the solution:

[1500(W) * 2 * 11(h) + 1000(W) * 6(h)] * 30 (days) = 1,350 kWh

Energy save in 1 month: 810 kWh = 37,5%

CO2 Emission save in 1 month: 40,5 kg (50g/kWh)

​

​

​

5. Compare

​

​

​

​

​

​

Electricity consumption for the solution: 0,12 kWh (86.4 kWh/month = 4,3 kg CO2/month)

6. Limitations and future works

Limitation:

• Instead of heater it was used a light bulb to simulate the heater behavior.
• The neural network has to be improved and integrated with the system.

Future works:

• Controlling of application in mobile phone.
• Implementation of the machine learning in real-time.
• Verify comfort levels by applying humidex level.

Creation application with options for users:

• scheduling;
• notifications;
• heating in advance

​

7. Demo

http://lutcodecamp.herokuapp.com/index.php

THREAD Protocol

Part 2:

Content

1. Introduction
2. Technology
3. Architecture
4. Case Study
5. Conclusion

Introduction

Introduction

• Thread is an IP based wireless networking protocol designed for low-power connected products in home automation space.
• Founded by 7 companies : Samsung, Google (Nest), Silicon Labs, Big Ass Fans, ARM, freescale and Yale security
• Support over 250 products per network
• Managed by your smartphone or tablet, and has a low latency period of less than 100 milliseconds for each typical interaction.

Introduction

​

• Thread features:
• Open standard protocol IPV6
• Simple for consumers to use
• Secure and encrypted (smartphone authentication scheme & AES encryption)
• Power-efficient
• No single point of failure
• Designed to support a wide variety of products for the home: appliances, access control, climate control, energy management, safety, and security

​

Technology

Overview of Thread stack

Architecture

Thread network topology and devices

Case study

Nest has created a program called "Works with Nest" using Thread that makes it possible for the Nest devices to securely interact with other "Works with Nest" devices.

Conclusion

• Thread can be a promising home automation technology
• Help device manufacturers focus on their product rather than on network protocols
• Fills a critical gap in the IoT ecosystem (low-power mesh networking solution, reliable, secure and scalable Internet connectivity)

References

Meier, A., C. Aragon, B. Hurwitz, D. Mujumdar, D. Perry, T. Peffer, M. Pritoni, “How People Actually

Use Thermostats”, in Proceedings of the 2010 ACEEE Summer Study on Energy Efficiency in

Buildings, American Council for an Energy Efficient Economy, 2010. accessed from

http://aceee.org/files/proceedings/2010/data/papers/1963.pdf

​

U.S. Department of Energy. 2015. Thermostats. http://energy.gov/energysaver/articles/thermosta

​