We are in the midst of an AI (Artificial Intelligence) explosion. Everything — from your refrigerator to your dog’s bowl — will become part of the AI neural network. By 2024 the AI market is expected to break the $500 billion mark.
However, AI has a problem – not in the algorithms themselves, but in how they are fed. AI needs a reliable, continuous, growing feed of live data from the IoT sensor networks surrounding us to make intelligent, real-time decisions. Capturing these data points requires a mesh of diverse, reliable sensors operating all the way from LAN to the edge.
So, what’s the problem?
As long as IoT sensor networks are powered by wires and batteries, AI is limited from reaching its full potential. What stands in the way is a dependable, flexible, scalable power source. Running wires to sensors is time-consuming, expensive and restrictive. Batteries are expensive, environmentally unfriendly, and there’s the labor and material expense of replacing them.
At Sensors Converge on June 22, 11:15 during the session, “Lightning Talks: Low-Power Design,” Dr. Charles Greene, chief operating and technical officer at Powercast, will discuss how wireless power can be used to unleash low-power wireless sensor networks (WSNs), enabling:
- at-scale expansion,
- untethered sensing in out-of-sight and hard-to-reach places, and
- elimination of disposable batteries and battery replacement issues.
Attendees will learn:
1) The pros and cons of various sources of wireless power:
● Energy harvesting – capturing available ambient energy from vibration, solar and wind
● Ultraviolet/infrared charging
● Inductive charging such as Qi
● Radio frequency (RF) charging
2) How RF wireless power in particular can cost effectively power maintenance-free WSNs. With one-to-many charging capabilities, one RF transmitter can reach and perpetually power many low-power sensors located up to 120 feet away. RFID readers used in warehouse and retail facilities today can even serve as the wireless power transmitters, meaning that the hardware infrastructure cost could be as little as the sensor devices themselves.
3) How to work within the rules of the US Federal Communications Commission (FCC), and devices that are best candidates. For example, FCC Part 15 imposes a 1W power limit. Low-power devices like sensors work well in this environment.
4) How RF wireless power networks (WPNs) work and how recent advancements like ultra-low-cost RF transmitters are enabling WPNs to proliferate.
5) How maintenance-free, battery-free, RF-transmitter-powered WSNs can contribute to an overall windfall by feeding AI with data. Examples:
● Stress/strain sensors installed inside bridges and roads can enable proactive monitoring and alerting to pre-empt infrastructure failures.
● Vibration/heat sensors installed inside wheels of trucks and trains can identify faulty wheel bearings and other issues to pre-empt train derailments and accidents.
● Fleet management systems can remotely monitor vehicle mechanical systems in real-time.
Rapid AI market expansion requires a reliable, continuous feed of live data from IoT sensor networks. However, reliance on wired and battery-powered sensor networks limits AI's potential. With recent advancements in ultra-low-cost RF transmitters, RF WPNs are becoming more accessible, enabling them to proliferate. By eliminating wires and batteries, low- power WSNs powered by wireless technology will undoubtedly strengthen the capabilities of AI.
At Sensors Converge 2023, Dr. Charles Green, chief operating and technical officer for Powercast, will address the topic, “Designing low-power sensor networks without wires and batteries strengthens AI” on Thursday, June 22, 11:15. Lightning Talks: Low Power Design Visit the Powercast booth, #840.