Cloud software for advanced building automation

AI Quick Summary

Control system upgrade vs retrofitA control system upgrade can be done much faster and at significantly lower costs compared to a retrofit of the building hull or the heating system.Why data predictive control?With data predictive control, the heating energy consumption can be reduced by 26-49% while improving comfort compared to state-of-industry control. Our method automatically learns the thermal behaviour of a building, uses weather forecasts and optimization to find optimal control inputs. ProblemBuildings are responsible for a third of the worlds’ energy consumption and CO2 emissions. Besidesexpensive retrofitting, their consumption can be reduced with efficient building control. However,buildings are often still operated with simple feedback control, especially in the residential domain.SolutionWith the help of algorithms based on a combination of physics-informed Machine Learning and mathematical optimization, we can reduce the heating and cooling energy consumption by 26-49% while increasing thermal comfort in a scalable way. Our method automatically learns the thermal behaviour of a building from a single week of measurement data, and uses weather forecasts and mathematical optimization to find optimal control inputs. Who are the customers and how will they profit from a solution?Building technology companies have developed smart thermostats and brought them to the market successfully. However, the underlying control technology is still the same as before and the now available data is not fully exploited yet. By integrating our cloud software service into their products, our customers can remain technology leader in the future, will be able to offer their customers energy efficient building automation, profit from service subscription fees, and will be prepared to enter into new and larger markets (i.e. demand response markets).How does your project idea affect energy savings and CO2 emissions?In experimental research, we have demonstrated 26-49% reduction in energy consumption for heating and cooling by using data-driven predictive control. All while the thermal comfort for the occupants was increased (i.e. 70% less cumulated comfort constraint violations). As almost 2/3 of the building stock in Switzerland is still equipped with fossil fuel based heating systems (i.e. oil and gas), these energy savings are directly related to savings in CO2 emissions. Moreover, a control systems upgrade can be done much faster and at significantely lower costs compared to a retrofit of the building hull or the heating system.Current status and previous activitiesSo far, viboo's data-driven control has been validated extensively in a research environment (NEST) during heating and cooling season. A first pilot project has been started recently with with the Danish thermostat manufacturer Danfoss. In this pilot project, 150 thermostatic radiator valves (TRV) in an office building were replaced through eTRVs and operated with viboo's data-driven predictive control. First evaluations confirm the results observed in previous experimental research. The next steps in the innovation process consist further pilot projects in different buildings types (e.g. floor heating systems) and the development of a market-ready cloud software.What are your planned work packages?WP1 - Development of cloud software solutionWP1.1: Elaborate suitable software architectureWP1.2: Evaluate different cloud infrastructure providersWP1.3: Develop cloud software according to findings of WP1.1 and WP1.2WP2 - Execution of pilot projectsWP2.1: Describe cases of further pilot projectsWP2.2: Identify suitable objects for the pilot projectsWP2.3: Install hardware and integrated systemWP2.4: Conduct and evaluated further pilot projectsHow can the Energy Lab help you?EnergyLab can help us to establish contacts with experts in software engineering and development, and cloud computing. Moreover, enlarging our network in the building technology industry will be extremly valuable as well. Exchange and workshops with software engineers and developers, and cloud computing specialists will boost the development of the commercial cloud software service. Getting in touch with building technology companies will potentially lead to new partnerships. Finally yet importantly, the EnergyLab InnoBooster funding will help us to conduct the pilot projects (support during implementation and supervision during execution) and finance the necessary IT consulting.

Project Idea Metadata

Project Idea Name: Cloud software for advanced building automation
Date: 2/28/2022 3:45:11 PM
Administrators:
- Benjamin Huber

Project Idea Description

Problem

Buildings are responsible for a third of the worlds’ energy consumption and CO2 emissions. Besides

expensive retrofitting, their consumption can be reduced with efficient building control. However,

buildings are often still operated with simple feedback control, especially in the residential domain.

Solution

With the help of algorithms based on a combination of physics-informed Machine Learning and mathematical optimization, we can reduce the heating and cooling energy consumption by 26-49% while increasing thermal comfort in a scalable way. Our method automatically learns the thermal behaviour of a building from a single week of measurement data, and uses weather forecasts and mathematical optimization to find optimal control inputs.

Who are the customers and how will they profit from a solution?

Building technology companies have developed smart thermostats and brought them to the market successfully. However, the underlying control technology is still the same as before and the now available data is not fully exploited yet. By integrating our cloud software service into their products, our customers can remain technology leader in the future, will be able to offer their customers energy efficient building automation, profit from service subscription fees, and will be prepared to enter into new and larger markets (i.e. demand response markets).

How does your project idea affect energy savings and CO2 emissions?

In experimental research, we have demonstrated 26-49% reduction in energy consumption for heating and cooling by using data-driven predictive control. All while the thermal comfort for the occupants was increased (i.e. 70% less cumulated comfort constraint violations). As almost 2/3 of the building stock in Switzerland is still equipped with fossil fuel based heating systems (i.e. oil and gas), these energy savings are directly related to savings in CO2 emissions. Moreover, a control systems upgrade can be done much faster and at significantely lower costs compared to a retrofit of the building hull or the heating system.

Current status and previous activities

So far, viboo's data-driven control has been validated extensively in a research environment (NEST) during heating and cooling season. A first pilot project has been started recently with with the Danish thermostat manufacturer Danfoss. In this pilot project, 150 thermostatic radiator valves (TRV) in an office building were replaced through eTRVs and operated with viboo's data-driven predictive control. First evaluations confirm the results observed in previous experimental research. The next steps in the innovation process consist further pilot projects in different buildings types (e.g. floor heating systems) and the development of a market-ready cloud software.

What are your planned work packages?

WP1 - Development of cloud software solution

WP1.1: Elaborate suitable software architecture
WP1.2: Evaluate different cloud infrastructure providers
WP1.3: Develop cloud software according to findings of WP1.1 and WP1.2

WP2 - Execution of pilot projects

WP2.1: Describe cases of further pilot projects
WP2.2: Identify suitable objects for the pilot projects
WP2.3: Install hardware and integrated system
WP2.4: Conduct and evaluated further pilot projects

How can the Energy Lab help you?

EnergyLab can help us to establish contacts with experts in software engineering and development, and cloud computing. Moreover, enlarging our network in the building technology industry will be extremly valuable as well. Exchange and workshops with software engineers and developers, and cloud computing specialists will boost the development of the commercial cloud software service. Getting in touch with building technology companies will potentially lead to new partnerships. Finally yet importantly, the EnergyLab InnoBooster funding will help us to conduct the pilot projects (support during implementation and supervision during execution) and finance the necessary IT consulting.

Control system upgrade vs retrofit

A control system upgrade can be done much faster and at significantly lower costs compared to a retrofit of the building hull or the heating system.

Why data predictive control?

With data predictive control, the heating energy consumption can be reduced by 26-49% while improving comfort compared to state-of-industry control. Our method automatically learns the thermal behaviour of a building, uses weather forecasts and optimization to find optimal control inputs.