Utilities can now rapidly solve previously intractable problems using next-generation AI/ML solutions on Google Cloud Platform to bridge the gap between insights and intelligence.
For technology professionals in utilities, there’s no more important topic right now than transformation and innovation powered by data and AI. Around the world, new developments, like real-time streaming analytics, wearables, and ‘paint-by-numbers’ build-your-own app tools, are enabling fast and incredible step changes in productivity – rewriting cost models.
- 42% of power company executives are exploring analytics, AI, and ML, and 35% will be exploring cloud computing in the next 1-2 years
- By 2022, 55% of utilities will use a core digital platform to automate, optimize, and orchestrate assets, processes, customers, and employees
- By 2023, utilities will have digitally connected 75% of their critical assets to predict and prevent equipment failure and extend asset life cycles
At the latest Executive Roundtable for Australian Utilities hosted by Wipro and Google, participants heard how utilities and energy companies use Google data and AI platforms to get transformative results in just weeks without needing an army of data scientists or ML engineers.
Using AppSheet to overcome digital transformation roadblocks
Like most utilities, a US-based Fortune 500 company was looking at ways to improve productivity and eliminate wasteful processes. Digital transformation of its core field operations had been stopped in its tracks by a heap of paper-based processes. Almost every internal process and workflow were tracked manually by error-prone documents that were time-consuming to complete. The company knew it needed apps to digitize these processes, but Paula – the primary app creator – had no coding experience.
Using Google’s AppSheet no-code app developer, Paula rapidly built ten apps (one in just 45 minutes) to drive paper-based processes out of their field operations. The apps, which have transformed productivity, include:
- Trouble Transformer Track – Linesmen now use this app to track transformers taken in and returned when a failed or damaged transformer is replaced. Rather than manually inputting serial numbers, they simply swipe a bar code. The fast, error-free process has delivered time and money savings and made tracking transformers easier.
- Circuit Inspection Tracker – This app replaced a spreadsheet (called ‘the beast’) to eliminate free form entries and collect GPS information on circuit defects. Users in the field can mark up pictures to provide additional context, improving speed and accuracy without compromising data quality.
Using Google Glass to empower frontline workers
ML and AI technology have now been integrated into Google Glass Enterprise Edition, offering a lightweight line-of-sight display giving frontline workers guidance in the field while performing tasks. Glass not only gives workers access to training videos, instructions, and checklists, it can connect with remote experts, allowing them to ‘see what they see,’ supporting real-time collaboration and troubleshooting. Research shows that when maintenance workers are fed images annotated with instructions, rather than having to keep switching back and forth between the task and a computer or tablet, productivity increases by 34%.
Imagine that result scaled out across all the technical activities currently performed in the field.
Using data science to help build a more sustainable energy market
A US-based Fortune 500 company is one of the world’s leading energy providers, operating out of 15 countries. Its mission is to accelerate a safer and greener energy future. Google is helping this organization accelerate its journey to AI maturity. Early use case examples include:
- Computer Vision for wind turbine inspection – The company has eight wind farms with between 50-300 turbines, requiring annual inspection across a vast spread of geography, including mountains. Drones take 30,000 images of each turbine, which previously required a highly trained engineer four weeks to review. The company used AutoML to automatically build the ML algorithms needed to eliminate half the images needing human review – training the solution using labeled data. Now 50% of the engineers’ time can focus on identifying damage and the right course of action to remediate it.
“We won’t reach the clean energy future without advanced tools like machine learning.”
AutoML supports an incredibly rapid development cycle. Google and Wipro have seen customers get proof of concepts up, demonstrating value in as little as 3-4 weeks.
- Stream analytics for turbine performance – Early in its AI strategy, the company used stream analytics solutions to get high volumes of telemetry data under control and available in real time. Today, it can identify and predict turbine gearbox and generator failures while using dashboards to give engineers a real-time view of what’s going on, helping to explore issues as they unfold.
- Smart analytics for market bidding – The company used Google tools to rapidly develop a model to support its analysts to make better decisions on when and at what price to buy energy – factoring in all the external signals that influence wholesale prices. The model predicts:
- Market conditions next day and week
- Competitor behavior
- Optimal price and quantity bidding strategy
With model and bid performance improving over time, predictions are capturing increasing value.
Using AI-based threat detection to reduce your attack surface
The recent uptick in ransomware attacks on critical infrastructure keeps pressure on utility cybersecurity teams. The Google approach to cybersecurity is to trust nothing. On Google Cloud, utilities have at their disposal an array of tools to reduce their attack surface and protect data and employees: from hardware tokens to AI-based threat detection, from device management to zero-trust networking solutions. Wipro is helping its utility clients use these tools to take a layered approach to security across users, access, data, and applications that help protect every click from malware, data loss, and fraud.
No matter what your primary public cloud provider is, utilities can still take advantage of Google data and AI platforms for all these use cases.
To discover how quickly you can use Google data and AI platforms to strip out layers of complexity and develop a proof of concept to solve your most pressing problem, contact us.