K–12 schools and districts are using artificial intelligence, specifically machine learning, to address a range of needs, from infrastructure efficiencies to targeted academic interventions.
Self-learning machines and intelligent algorithms can detect the signs of students vaping on campus or spikes of noise that might indicate a violent incident. AI-driven innovations can collect and analyze data on HVAC usage to help administrators identify inefficiencies. And those are just a few examples. Such evolving technologies promise plenty of benefits for K–12 education.
So, what is machine learning? “Machine learning algorithms use statistics to find patterns in massive amounts of data,” according to the MIT Technology Review.
A key benefit of intelligent algorithms: detecting patterns in vast data sets that frustrate human efforts. Advanced machine learning tools now leverage human-inspired “deep neural networks” to deliver both pattern recognition and behavioral prediction, while AI solutions are designed to mimic human decision-making based on available data.
In K–12 education, machine learning tools enable collating and correlating student performance, and then identifying key indicators that suggest the need for specific teacher or administrative support. Administrators also have to navigate the privacy and security concerns surrounding AI-driven deployments — a growing challenge as the use of Big Data in education becomes more commonplace and districts’ fleets of digitally connected classroom devices expand.
There’s also more to machine learning than classroom data collection.
MORE FROM EDTECH: How K–12 Schools have adopted artificial intelligence.
Leveraging Machine Learning to Identify IT Patterns
Many institutions now use machine learning to search for patterns and sift through operational IT data, says Mohan Rajagopalan, senior director of product management for Splunk. Doing so, he says, empowers them to detect “anomalies, such as deviation from past behaviors indicating machine or network failures, or unusual changes in access patterns indicating potential security issues that may arise,” allowing IT staff to forecast usage trends and assist in capacity planning.
That data analysis is beneficial to K–12 schools running on last-generation network technology while simultaneously managing one-to-one computing initiatives. Having the ability to predict potential downtime and understand student use trends can help administrators more effectively track technology spending and security. Machine learning leaders such as Splunk have already helped school districts prevent network outages and reduce their mean time to investigate and repair IT issues.
MORE FROM EDTECH: Teachers are turning to AI solutions for assistance.
Turning to AI to Help Shore Up Staffing
Education has a tech talent shortage. That’s no surprise: K–12 schools often can’t offer competitive salaries, and many districts are located outside of large urban areas, making it harder to recruit from an already-limited talent pool. The result? Local teams are on the hook to run enterprise-scale networks with skeleton crews.
Institutions can leverage intelligent algorithms “to supplement and augment human operators,” Rajagopalan notes. For schools, these solutions offer a way to do more with less by enhancing the efficacy of smaller IT teams tasked with servicing technology solutions at scale, implementing data-first security features that prioritize student privacy and supporting both in-house and BYOD deployments.
Machine learning integration offers key IT infrastructure benefits, Rajagopalan says, including:
- Automatic detection — Based on existing use and behavior patterns, algorithms can detect problems across the network.
- Simplified investigation — Machine learning tools make it easier to identify root causes of identified network problems.
- Active triage — When paired with other AI solutions, machine learning empowers automatic triage of specific events, such as application misuse or unauthorized network access.
- Alert routing — For events that can’t be automatically triaged, reports are generated and routed to IT admins for further investigation.
- Enhanced speed — Splunk now sees customers “leveraging machine learning for optimizing security orchestration, automation and responses to execute actions in seconds, not hours,” Rajagopalan says.
MORE FROM EDTECH: Assessment innovation in K–12 levels the playing field for students.
Using Algorithms to Foil Facilities Failures
Effective school environments extend past classrooms, teachers and learning technologies to the basic building infrastructure. For example, sudden HVAC failure could cause building temperatures to plunge or skyrocket, forcing temporary closures or class relocation. Inefficient devices can also negatively impact school budgets if districts overspend on maintenance or replacement, draining funds that could be used for new computing technologies such as virtual reality assets or cloud-based assessments.
Technology companies such as Microsoft already are leveraging machine learning to reduce climate control costs and improve employee comfort. But recent research suggests schools — by virtue of their not-for-profit approach — often overlook cost-effective investments in energy efficiency.
The sheer amount of data that facility control systems and sensors generate provides the necessary foundation for machine learning to automatically look for failures and resolution, Rajagopalan says. He points to the example of machine failure due to overheating: If school facility managers received alerts of air conditioning units overheating based on current ambient temperature and usage patterns, they could temporarily shut down the units for repairs, reducing the need for costly replacement.
Finding the Right Machine for the Job
Despite the “mind-boggling” potential for machine learning, Rajagopalan says, “the recipe for success lies not in developing more technology but in being able to successfully align technology with specific use cases and user needs.”
For K–12 schools, that means the application of machine learning and AI isn’t about speed or scale, but specificity. From identifying IT patterns to bridging the tech talent shortage or avoiding costly failures, machine learning applied to solve specific challenges can help school districts maximize their digital strategy investments.
