News & Events
5 Predictive Technologies That Will Make Your Life Better - Popular Mechanics
When I walked into my room at the InterContinental Hotel in San Francisco recently, I noticed the fans start to gently heat up the room. (It was an unusually cold day.) A motion detector had noticed my entry into the room. But there was something else much more impressive occurring deep in the lower recesses of the hotel. A new system currently in the test phase, called SCIwatch, was monitoring energy use in the entire hotel, looking for anomalies.
SCIwatch can tell when a sensor in one small part of the building is malfunctioning. The system can monitor temperatures and generate a stock of historical data for any part of the hotel; then can tell it if there will be too much of an energy demand for a given week. And much of this analysis occurs in real-time: SCIwatch monitors thousands of systems, sensors, and controllers within the hotel, feeding data to a core server about every 5 to 15 minutes, depending on the end point. The sooner the system detects an error, the sooner engineers at the hotel can act on that problem. Essentially, they can use the system to predict problems before a hotel guest can even report the problem.
The hotel's main goal is to predict energy use throughout the year to control future heating and cooling costs. The historical data could be used to arrange a better agreement with a local utility provider. It could also help the hotel keep its customers. In most cases, the time you find out that a hotel's a/c system is not working properly is when you wake up at night in a cold sweat. And that means you'll probably never stay there again.
In the future, a monitoring system in your home or office could perform a similar function. The Nest thermostat, for example, released last year, learns your daily routine and can predict when to turn down the heat because no one will be in the house. Once home appliances and even the backyard grill are connected to a home server, a future energy-monitoring system could predict exactly what you want to do (and when) and adjust services accordingly.