It was almost a year ago when we concluded a project named “The Smart Campus Simulation Tool”. We are looking to release the simulation tool to open source. In this post, we wanted to explore the problem context which informed our design.
To us a Smart Campus represented a socio-technical system that would be “malleable” enough for us to achieve our objectives. We approached it to be a socio-technical system, the technology (the adaptive sensor based control system) has to work with the social context of an academic institute. At the end of the day, people have to accept and be willing to make changes to their lifestyles.
We wanted to look at the issue of electricity consumption for the IIIT-Bangalore. The institute had invested in a fair amount of energy saving equipment such as solar panels and more efficient water heating systems. But, they were not clear about the eventual savings in energy, the electricity consumption patterns or if there was a strategy to reduce the overall consumption in the campus.
Consumption of electricity is a difficult notion to comprehend and convey. For example, when a switch is thrown, does one wonder where the electricity is generated from? It may so happen that a forest is being cleared in Chattisgarh so that you may be able to spend an extra hour on Xbox. Furthermore, we have an inherent expectation (if you grow up with some privilege,) that electricity “has” to flow if a switch is turned on. People who have no access to electricity are vulnerable in many ways to the extent that their social mobility may suffer due to lack of electricity. People who have intermittent access or pay huge bills are also cautions about consumption. Nevertheless, we seldom question the source of generation.
Causal relationships like the one above between your consumption and environmental degradation are common and are uncomfortable (but true). Such examples try to guilt you into changing your consumption behaviour. However, it is not an easy to make lifestyle changes nor is it easy to ponder on the utility before doing everyday mundane tasks. Responsible use of electricity requires changes to behavioural and cultural practices as well as upgrades to the technical systems around us. Looking at both social and technological aspects was the cornerstone of our approach.
We tried to look at the campus as a location which enables different people to achieve their academic goals. People in the campus perform various activities that allow them to achieve this goal. We looked at activities that consumed electricity. We then developed a simulation tool that assumed the use of sensor -based control and behavioural modification to try and check if a technology-assisted behavioural change was possible. The results of the simulation would be the base to design a serious game. The game in conjunction with sensor-based control systems would address both social and technological aspects of the issue.
Our simulation mainly consists of:
- a model for generating activities (explanation for what this activity means below) for various actors present on the campus,
- an agent based model for minimising electricity usage while keeping the comfort level of individuals at an acceptable level.
We define an activity as any action that an individual takes during the course of one’s day in the campus. A good way to model an activity is to collect detailed information using “energy dairies”. As a small academic institute, the campus had limited types of actors. We therefore chose to use a survey-based approach to collect information on daily routines. We conducted a survey to understand various daily routines for all the individuals on the campus. We also conducted interviews with some of the administrative and housekeeping staff. We used this information to create a model for the generation of activities for various actors on the campus.
To model the “smart” systems of the campus, we created a control mechanism based on autonomous agents trying to collectively bring down the electricity consumption of the campus while keeping track of inhabitant’s comfort levels. We modelled the rooms and work areas as the autonomous agents. Each such agent was responsible for the operation of various devices that would consume electricity. It was then tasked with the objectives of minimising usage of certain devices by:
- negotiating the electricity consumption with other rooms (agents).
- Directing uses to use more common areas.
- Restricting when possible, the use of high power consumption devices such as air-conditioners and elevators.
In all of the above cases the it is assumed that the individual can override the agents, thus, keeping the human at the centre of the system. (This also allows us to collect information on what sort of activities will not be compromised in the name of energy savings. ) However, a denial from the system to allow the operation of devices resulted in a decrease in the satisfaction of the inhabitants. The agents were asked to minimise the use of electricity with as little discomfort as possible for the inhabitants.
Once the models were ready we created a simulation tool and calibrated it based on the data collected by the campus for over a year on a daily basis. We could then play out scenarios such as:
- What happens when we want to aggressively minimise consumption
- or, what happens when the comfort for the inhabitant is paramount and
- finally, what happens when we set a electricity consumption target for ourselves?
It was very interesting for us to see the results and present it to the inhabitants of the campus. We are now trying to work with students to create and deploy the sensor systems at the campus. We see a potential for extending this tool to include larger spatial/network levels such as a neighbourhood or a set of neighbourhoods as opposed to a campus. We are also looking at including multiple sources of electricity, given that decentralised power and micro-grids can become popular. Furthermore, we are also exploring the possibility to include other resources such as water consumption and sewage as well into the analysis. For a more detailed description to the tool and to some other people doing similar work please refer to our paper “Krishna, Harsha, Onkar Hoysala, Krishna G. Murali, Bharath M. Palavalli, and Eswaran Subrahmanian. “Modelling technology, policy and behaviour to manage electricity consumption.” In Humanitarian Technology Conference (R10-HTC), 2014 IEEE Region 10, pp. 40-45. IEEE, 2014.”. We hope to produce and publish more results soon. In the meantime please free to check our tool at:The Smart Campus Simulation Tool