One of the promises of the Internet of Things (IoT) is its ability to help companies improve their supply chains by gathering data that can be used to answer such questions as when goods will arrive and the quickest and safest routes for transporting goods.
Savi, a provider of sensor technology and sensor analytics solutions, has added new predictive and prescriptive analytics-based scenarios to its Savi Insight software, a software-as-a-service (SaaS) solution that captures data from sensors and other sources, correlates multiple variables including time, temperature and location, and finally applies logic that the company says turns data into actionable intelligence.
"Savi Insight allows our clients to quickly measure, assess and predict performance based on their own data and to rapidly benefit from a growing list of pre-packaged scenarios that address their top business challenges," said Bill Clark, Savi's president and CEO, in a statement.
The idea is to help companies spot opportunities, avoid problems and benefit from prescriptive analytics that recommend actions to improve future outcomes.
For example, Clark said, some Savi customers use sensors to measure the physical environment and location of remote high-value assets." After analyzing millions of sensor readings our algorithms can derive specific locations of interest and score them based on their propensity for crime," he said.
Another example: Savi Insight can capture real-time sensor information, compare it to historical models and score the likelihood of a shipment being late.
A key challenge in analyzing sensor data, Clark said, is capturing and cleansing the data before it can be analyzed. Sensors are often deployed in harsh environments, and they typically use battery power, low-cost hardware and wireless communications. "These factors all contribute to frequently missed, duplicate or unreliable readings," he said.
To help mitigate these issues, Savi uses an intelligent IoT adapter to capture any sensor data format -- structured, semi-structured, unstructured -- and transform it into a common Sensor Message Format (SMF). A complex event processing engine then uses pattern detection and a rule base to identify and resolve duplicate messages, missing messages and other data quality issues before the data is analyzed.
Savi Insight is also "tag agnostic" and thus can ingest and correlate data from nearly any source, including RFID tags, barcodes, ERP systems or even Twitter.
"As customers become more aware of the IoT they are starting to understand the value of combining other sources of data with sensor-generated data. Open source data such as weather forecasts or Twitter feeds, as well as enterprise data from corporate data warehouses, can be combined with sensor-based data to help put into context the exposure to risk or opportunity to enhance performance," Clark said.
Savi serves nearly 600 clients, including the United States Department of Defense.