Four weeks have flown by, wow.
So...I have been intensively working on market research and outreach for the NI4AI project. Working with the team, I have drawn up a list of utility companies with smart grids in the US. And I have been looking into innovative programs run by utilities and institutions; there are a few. We are hoping to find collaborators who would share their data on the NI4AI platform.
Sean, the CEO of Ping Things, explains at a demo for NI4AI that “Artificial Intelligence is a Fundamental Technological Advancement.” This advancement is almost like the jump from using trigonometric tables to calculators. In addition to the obvious merit of time and cost saving, collecting high frequency and time series data can not only pick up minor meter issues but also possibly prevent power outage or even wildfires. One of the main objectives for this project, which also makes NI4AI innovative, is to make data ACCESSIBLE to the public.
In the Smart Grid System Report in 2018, the Department of Energy states that “we have witnessed the accelerated deployment of technologies meant to improve the reliability and efficiency of utility operations, including the deployment of systems and practices to better engage utility customers in the management of energy.” Benefits of modernized grids are not just encouraged but it has also come to the point that it is not possible to meet consumers’ demands and keep the power resilience. Though the level of regulations vary by state, the map shows the AMI deployment progress status per State as of 2016.
Who is already utilizing the collected data?
This is one of my research objectives. Some large investor-owned utilities tend to have been exploring analytical tools; some of them collaborate with startups like Ping Things and universities. A lot of small, municipal and cooperative utilities have or are in the process of deploying smart sensors, but data is collected mainly for billing or reduction on power usage for customers.
The issue is that it is difficult or appears daunting for small utilities, which are often run by a small workforce, to utilize data or hire data scientists. The irony is, a lot of small utilities serve rural areas especially in wildfire risk areas and can benefit a lot from time series data.
The NI4AI platform can fill this gap among utilities and potentially take utility management to the next level! Consumers can benefit hugely if the whole industry utilizes data to have better predictions.
(by Sean Murphy CEO of Ping Things)
What does it take to share data?
I am currently working on promotion to answer these common concerns:
- Data Analysis Knowledge, skills and resources required?
Utilities do not necessarily have huge resources or the requisite number of data scientists to participate in data analysis. Using platforms like NI4AI, makes it simple to analyze the data. Simply play around with the web-based plotter where you can zoom in and out of the point you are interested in and explore any irregular streams. You can also export a CSV file or export a jupyter hub for local analysis. Some coding samples, for anyone to use, are also available.
- Secure in the Cloud?
A lot of utilities hesitate to share data for security concerns. It is much safer than saving data locally or some hard disk as cloud storage has more professional security in place. Using Cloud also enables utilities to access cloud computing; does not limit to the ability of users’ machines.
- Why publish in public?
Publicly-available datasets can lead to infinite possibilities. Academia and students can use data to research, for example, innovative new technologies, wildfire indications and train future engineers. Data scientists or software engineers can play with data to build tools.
So my search continues…
Next week, I will be interviewing innovative program leaders from municipal utilities. Stay tuned!