Microsoft cloud tool makes data analytics apps simpler to build

Wed, 2014-09-24 06:00 -- SCC Staff

Machine learning technology is expected to spawn a new generation of smart cities solutions that take action based on analysis of data collected from sensors and other sources.  And the launch of Azure Machine Learning (ML) from Council Lead Partner Microsoft could do much to accelerate the process.  

Reviewers assert that the cloud-based Azure ML platform lowers the entry bar for solution developers wanting to build data-driven applications to predict, forecast and change future outcomes. The web service sharply decreases the complexity and cost of developing machine learning systems.

“Machine learning today is usually self-managed and on premises, requiring the training and expertise of data scientists,” explains Joseph Sirosh, corporate VP of Microsoft Machine Learning.

“However, data scientists are in short supply, commercial software licenses can be expensive and popular programming languages for statistical computing have a steep learning curve. Even if a business could overcome these hurdles, deploying new machine learning models in production systems often requires months of engineering investment.”

Making campus building efficient and comfortable

Several institutions and Microsoft partners have been testing the Azure ML platform, which is still in preview. Among them is the Center for Building Performance and Diagnostics at Carnegie Mellon University.

The center’s researchers wanted to create a system to boost the overall energy performance of their campus buildings and ensure that occupants remain comfortable. They identified real-time factors that affect occupant satisfaction as well as weather forecast data to help predict cooling or heating requirements. Council Associate Partner OSIsoft helped with the data collection and to develop a system to predict energy consumption across buildings.

The project team used Azure ML to analyze the data and come up with an integrated hardware and software solution. The university has seen energy reductions up to 30% in some buildings after the system was deployed.

OSIsoft Director of Research and Innovation Gregg Le Blanc said that using Azure ML provided “the ability to predict the implications of different actions based on acquired data in the PI System and store the data within the PI System for exploratory investigations. The result will be the ability for our customers to predict more and test less.”

Machine learning is already the driving force in a number of today’s familiar analytical applications, including search engines, online product recommendations, credit card fraud prevention systems, GPS traffic directions and mobile phone personal assistants. For a more technical dive into machine learning, check out the two-part series "Machine Learning and Cognitive Systems: The Next Evolution of Enterprise Intelligence" at Innovation Insights.

Related articles …

Smart buildings: Predicting your buildings' energy needs

What IBM's 5 in 5 predictions say about how city life (and city hall) will change