FAQ - AWS Prescriptive Guidance


Can I benefit from visualization without implementing an RL solution?

Yes. Techniques such as hot spot detection, cohort comparison, historical trending, and other descriptive analytics can provide value without the need to invest in predictive and prescriptive analytics. For more information about these techniques, see the links in the Resources section.

How many data points do I need to take advantage of RL for energy optimization?

We recommend that you collect at least one year of historical data in order to capture trends from the previous four seasons and to take the highs and lows of the entire year into account.

How should I start an RL project?

We recommend starting with a single site that has average technical capabilities yet meets the suggested requirements in the Best practices section. You can then define a roadmap with a phased approach to scale out to other sites. This approach ensures project feasibility and provides long-term scale-out potential.

Which industries can benefit from this strategy?

Organizations across a variety of industry sectors can benefit from RL-based energy optimization. These include manufacturing (discrete and process), oil and gas, renewables, power and utilities, healthcare and life sciences, agriculture, retail, and packaging companies that want to save operational costs by optimizing the consumption of energy and other resources (such as water and gas) for commercial equipment. This strategy can help these companies save resources and reduce operational costs, carbon emissions, raw material consumption, and waste disposal.

What are some of the successful use cases?

Proven use cases include optimizing chiller usage in pharmaceutical manufacturing, optimizing mill usage in raw material processing, optimizing nutrient dosage in agriculture applications, and optimizing HVAC energy consumption in Amazon fulfillment centers. To read about these, see the links in the Resources section.