Machine Learning for Identifying Energy Use Patterns for My School District

The average school in the U.S. is 40-years-old, reports the United States Department of Energy. Buildings of this age, as well as those that are much older, are not energy efficient. As explained by Fiona Burlig of Forbes, K-12 schools spend a cumulative $8 billion annually on energy, and given that Cenergistic-partnering schools reduce energy costs by at least 25 percent, that could amount to a savings of $2 billion if every school in the country were to partner with Cenergistic. With energy costs increasing across the country, this expense will grow. School districts need to understand the role of new technology, including machine learning, to answer the question “how can identifying energy use patterns for my school district affect energy spend?” Since we are focused on the school facility manager, let’s approach this topic with key questions a facility manager might ask after the CFO implements budget constraints.

How Can My District Use Machine Learning Without Being a Programming or Data Expert?

Machine learning refers to the use of historical and comparable data to self-optimize and accommodates changes within an existing algorithm. In other words, machine learning allows a system to learn from itself to achieve the desired outcome. In this case, less-energy use is the desired outcome, but machine learning is an advanced topic, requiring knowledge of big data and analytics, so it is shelved in favor of more straightforward solutions.

The other side of the problem with using machine learning goes back to issues with the realization of cost savings for energy-efficiency improvements. While the use of machine learning has been studied for its ability to improve decision-making processes, the best developments in the world cannot correct wasteful energy use behaviors. The use of machine learning for identifying energy use patterns for school districts can reduce the impact of destructive behaviors. To maximize savings from machine learning, facilities managers combine technology and on-site energy specialists.

How Can Machine Learning Improve My Understanding of Energy Use Patterns for My School District?

Energy management decisions are not merely about reducing energy use during vacant hours. They reflect the need to reduce energy consumption through conservation. Machine learning transforms raw data into actionable insights designed for a specific need. If students or staff in one part of a facility continued to leave lights on or run HVAC systems beyond recommendation, machine learning could be used to define the steps necessary to achieve savings.

Machine learning is not the same as full automation. Facility managers are still necessary to make changes to the systems and act on information. More importantly, automation relies on pre-set variables, so poor data and application of data can have a disastrous effect on automation. Therefore, understanding of energy use patterns decreases. Ultimately, using machine learning will give facilities managers a crash-course in energy-use patterns occurring and how to change habits to save more money.

How Does Knowing Energy Use Patterns for My School District Affect Savings and Funding?

Modern technology can help the facility manager track the runtime, energy use, maintenance needs, life expectancy, vacancy runtime, and countless other factors for individual assets. Combining this information is more valuable than gaining asset-level data itself. It allows a facility manager to benchmark the performance of a given asset against the overarching performance in the facility, and in the same fashion, the application of machine learning for identifying energy use patterns through an appropriate benchmarking process can increase accuracy in determining cost savings through improvement.

A common factor for facilities managers in schools is lack of funding, but funding for energy management may be available through government programs that incent energy efficiency in school districts. Browse the hundreds of “State Incentives for Renewables & Efficiency” featured at DSIRE, the most comprehensive source of information on incentives and policies that support renewable energy and energy efficiency in the United States. However, school district officials must know that poor understanding of energy use patterns in the first place, without machine learning, will lead to a lower return on investment from a facility asset upgrade than expected and continued eligibility in many, if not all, of the states' initiatives listed in the database, require schools to demonstrate returns.  

How Can I Put the Power of Machine Learning to Work in Energy Use Patterns in My School?

This is the most natural question to answer. The facility manager needs to begin by taking stock of current energy use patterns, namely utility costs. Determining annual energy expenses is the simplest way to review energy costs. Then, facilities managers will need to take a step further, leveraging technology and available resources, such as those offered by Cenergistic, to gain the insight into facility condition, asset performance and potential savings through machine learning. If your school is ready to kickstart your energy management program with the right data and the help of an on-site data specialist, contact Cenergistic online or by calling 1-855-798-7779 today.

Submit a comment: