Keeping tabs on multiple electrical devices for making sure that which one is failing or running properly can be quite challenging. MIT’s NILM (non-intrusive load monitoring) system has been designed to help by making use of sensors and a computer dashboard for tracking the devices’ status in environments including ships, high-rises, or factories.
The core of the NILM system has a sensor that is mounted on the exterior of a single wire within the electrical circuit. There is no need to splice or cut the wire. The sensor will begin monitoring the current that is flowing through the wire and will keep tabs on different devices that are being run on the circuit. How does it do that? It makes use of telltale fluctuations in the current, which are synonymous with device switching on or off.
By utilizing these readings, the system can ascertain when and if each device is running and can also determine if the device is using more current than is normal. This particular feature enables the NILM to figure out if a device has become defective.
The data obtained is sent to a central computer that features dials on an onscreen dashboard for each device. If the needle is in the green zone, all is well. However, you need to check up on the circuit if the needle is in the yellow or red zone.
The NILM system has already undergone testing last year on the Coast Guard cutter ship Spencer. Two of the NILM sensors were utilized for keeping checks on 20 unique devices. The NILM sensors were able to identify that a certain ‘jacket water heater’ located on one of the diesel engines was using unusually high amounts of power. Investigations revealed that the heater had undergone severe corrosion and had broken insulation. If this was not identified when it was, the heater would have caused an electrical fire.
The NILM sensor that had identified the issue did so via a hard-wired connection. However, it should be noted that the NILM system can also work wirelessly. The NILM system doesn’t require any internet connection; thus the communications network can’t be hacked by third parties. The research is being led by Professor Steven Leeb, and a paper on it has been published in the journal IEEE Transactions on Industrial Informatics.