Today’s manufacturers are eager to explore opportunities for enhanced energy efficiency in their factories. Besides saving money, these solutions could align with their sustainability goals and reduce wasted resources. Many emerging options combine artificial intelligence and the Internet of Things. What can AI and IoT technologies do for those who strategically apply them?
Achieving enhanced energy efficiency through better monitoring
Decision-makers may set broad goals to use less energy but need help refining that aim. Fortunately, AI and IoT solutions can provide granular details about energy trends and recent progress.
In one case, Japanese pharmaceutical executives set a goal to achieve a 1,600-ton CO2 emissions reduction by 2025. A significant part of their strategy involved an automated building management system that uses IoT sensors and AI to help building managers make better energy-efficiency decisions. Additionally, the system uses building data to establish benchmarks and automatically adjust power usage as required. That functionality reduces the need for manual oversight.
However, the technology also includes a remote monitoring component that allows authorized users to visualize and control manufacturing processes in real time. The AI and IoT technology is also highly specific, operating individual plug outlets and loads or the lights in particular areas.
Connecting these high-tech energy-efficiency technologies and tools to building management systems is a practical way to streamline operations and achieve progress in ways people might not otherwise consider. Perhaps the data shows a factory uses more electricity than necessary because an automatic door stays open longer than required, activating the climate-control system more than expected. Changing the timing through the building management platform could make a meaningful difference.
Preventing unknown resource waste with AI and IoT
Some manufacturing leaders realize unaddressed wastage has prevented them from achieving their energy efficiency goals. Sometimes, such issues occur due to simple oversights, such as if someone leaves the factory’s lights on for the weekend, and no one notices until operations resume the following Monday. However, AI analytics and IoT sensors on light fixtures could tackle that issue, automatically switching them off after not detecting motion in the area for a predetermined time.
Unoptimized processes also cause preventable energy waste. It may seem like everything is running smoothly until the electricity bill comes next month, and it is much higher than usual.
Factory equipment is a common culprit for such inefficiencies. For example, statistics suggest people can reduce air compressor energy usage by 35% or more by choosing models with variable-speed drives. Additionally, leaders can complement those options with centralized control systems that alert them to abnormalities or allow setting the equipment to operate only when needed.
Another example combines AI and IoT technology to offer predictive maintenance insights for industrial air compressors. Various sensors collect data on humidity, flow, temperature, pressure and other parameters, then use that information to learn the equipment’s expected performance and flag deviations. Additionally, a current transmitter and real-time data streams notify technicians when compressors work harder than they should, indicating potential problems that could waste energy and cause other adverse effects.
At one plant that uses this platform, the data showed a compressor running at a loss and consuming too much energy. However, plant managers identified the problem by reviewing the data and found a straightforward fix. People can only begin to thoroughly address unintended energy use after knowing problems exist. Then, they can use the associated information to target the identified reasons for the excessive energy use.
Using AI to reduce IoT device energy usage
A manufacturer may have hundreds or thousands of IoT devices at a single facility. Their collective energy use could be substantial, especially since many low-power options are still in development or not always suitable for the industry’s needs.
However, a research team made an industrial automation breakthrough for IoT devices that relies on backscatter communication. In this approach, the connected products alter the signals’ load impedance, reflecting and modulating the signals instead of generating them. This method has notable potential, but previous efforts elsewhere showed discrepancies between expected results based on simulations and real-world outcomes. Those typically caused differences in the modeled and actual reflection coefficients.
The group overcame those challenges with an artificial neural network trained through transfer learning. That technique involves algorithms that enhance their performance on specific tasks based on how they previously handled related ones. The AI-driven training method resulted in a deviation of only 0.81% between the simulated and real reflection coefficients.
This significantly better accuracy allowed the researchers to enable energy-efficient data transmission between IoT devices. They also believe their work will allow reliable and efficient performance, based on how the team achieved a 9.35% error vector magnitude. Overall, the researchers made their connected devices 40% more efficient than conventional options by using AI.
While this work is in the early stages and not yet part of commercial IoT devices, it should provide other interested parties with valuable insights and inspiration for how they could use AI algorithms to make IoT devices more energy-efficient. Such successes could also impact other areas. For example, these researchers see potential applications for their innovation extend to smart cities and public service applications, suggesting people in other industries should consider adopting similar possibilities.
Applying IoT and AI Technologies to improve quality control
Manufacturers with high rework rates may use more energy than intended because of the additional machine running time needed to correct flaws in products that do not pass quality control checks. However, using the IoT and AI to find process problems sooner could increase instances of products made right the first time.
Getting the best results requires identifying the most significant aspects that cause undesirable results and setting associated goals. For example, a company might use AI-driven machine-vision algorithms to improve its non-destructive testing for automotive parts. Similarly, IoT sensors on critical equipment could warn them of misaligned components that will cause quality control shortcomings. Reliable and real-time data from IoT hardware and AI algorithms can immediately tell manufacturing leaders about abnormalities, giving them the insights needed to thoroughly address them.
These technologies can also assist if plant executives realize inconsistencies prevent them from meeting quality control standards. AI and IoT sensors excel in such situations due to their data collection and analysis capabilities.
In one example, researchers used data analysis to make concrete approximately 30% stronger by altering the mixture to improve performance without additives. A similarly targeted approach could bring enhanced energy efficiency if manufacturers determine which factors cause inconsistent quality in some products, resulting in more overall work to produce the necessary volumes.
Enhanced energy efficiency is possible
Although manufacturers must take time to find the most appropriate ways to use AI and IoT technologies, these examples show how both can be instrumental in meeting energy efficiency goals. Choosing metrics and tracking them throughout implementation periods and trials is an excellent way to learn which applications work best.
