Your IoT projects cover more than a mere few sensors and dashboards. Using the IBM Watson IoT™ Platform you will have many more possibilities at your disposal, such as integrating weather data, Natural Language Processing and cognitive and predicting analyses.
The Weather Company is specialised in the delivery of weather data. Also in The Netherlands they have a network of data sources. With a few clicks they can easily be added to your other IoT-data, making dependencies and relations become insightful.
Did you know that a study has shown that The Weather Company provides the most accurate weather data in the world? Within the Watson IoT-platform you can instantly make use of this weather information and, as such, gain more insight in how the weather affects your maintenance processes.
Natural Language processing
Due to Natural Language Processing (NLP) users are now able to communicate with systems and devices by use of simple human language. Natural Language Processing helps solutions to comprehend human language by correlating the language to other data sources for the relevant context. For example, an engineer, operating a machine, may notice an unusual vibration. Subsequently he asks the system: “What causes vibration?”. By use of NLP and other sensor data, the system automatically links these words to the probable connotation and intention. The machine then studies recent maintenance in order to identify the most plausible source of the vibration and to advise an action to resolve it.
Machine Learning automates data processing and continuously checks new data and user interactions in order to classify data and results, based on the priorities entered into the system.
Machine Learning can be used for monitoring the device and sensor data and automatically comprehend the current conditions as well as the expected trends. In case of expected problems it can provide suggestions for taking the appropriate actions..
For example, the platform can monitor incoming data of fleet equipment in order to familiarise with normal as well as abnormal conditions, including environmental and production processes, which are often unique for every device. Machine Learning helps to understand these differences and configures the system to actively check the unique conditions of each device.
Video and Image analysis
Using video- and image analysis it is possible to monitor the unstructured data of video feeds and snapshots to identify scenes and patterns. This knowledge can be combined with machine data for a better understanding of past events and future situations. For example, security cameras can detect the presence of a fork lift truck in a confined area and send an alert to the system. Suddenly, three days later, there is reduced activity in that area. The two incidents are correlated in the system to identify a possible collision with the fork lift truck that the video maybe didn’t clearly indicate.
Text Analytics makes it possible to link unstructured text information, including transcripts of customer control centres, logbooks of service experts, blog comments and tweets, to detect correlations and patterns in this massive quantity of data. For instance, phrases reported by unstructured channels – such as, “my brakes make a sound” or “the pedal feels dirty” – can be linked in order to identify possible problems with a certain car brand and model.
IoT Network Support
Part of the success of the IoT application is due to the availability of new networks that transmit data at a low bandwidth and low energy consumption.
Low Power, Wide Area Networks (LPWAN) are perfect for connecting devices that transmit small amounts of data over a long distance. This technology enables these devices to submit their data using very little energy, which makes battery powered devices last longer, even years, as some cases have proven. The advantage is that sensors can operate at remote locations, where there are poor facilities, or even none at all. This “new” data gives new insights and provides new possibilities.