The aim is to predict machine failures in by analysing information from sensors in real-time.
“AI is becoming a key enabler for predictive maintenance and performance improvement, because of its cognitive abilities such as learning, reasoning and problem-solving. Rohm and A*Star will develop an AI chip that is capable of processing and analysing data as soon as they are received by a sensor node,” said Rohm.
The reason for developing a chip to move intelligence to the sensor, rather than using a central server, is restricted data bandwidth on the wireless sensor networks expected in these applications. Some of the raw data – vibration for example – has considerable bandwidth, which local processing can reduce.
“As the number of sensors increases in the future, the communication technology for wireless sensor networks would face bandwidth constraints, and be unable to expeditiously transmit the increasingly large sensor data to the computer server,” said Rohm.
In the partnership, Rohm is supplying AI analytical algorithms, while A*Star’s Institute of Microelectronics (IME) has capabilities in ultra-low power mixed-signal chips.
Both organisations also have experience in analog ue computation circuits.
“The research collaboration will enable the developed chip to filter volumes of data across multiple sensors, and analyse complex data patterns in real-time,” said Rohm. This novel AI chip is expected to perform significantly faster than the conventional method for predictive maintenance, as well as reduce power consumption. Rohm has plans for the chip to be compatible with wireless technologies such as Wi-SUN, and EnOcean 2, and to incorporate the chip into its proprietary sensor nodes and wireless modules. ”