Why Software is More Important Than Sensors in the Internet of Things

As the Internet of Things slowly becomes a commercial reality, led by industries such as food and logistics, the underlying technologies (RFID, sensors, QR codes and more) become less important than what is done with the data. As Bo Begole, Principal Scientist and Manager of Ubiquitous Computing at PARC, put it to me recently: "the algorithms are more interesting now than the sensors."

Begole’s group at PARC (Palo Alto Research Center, a subsidiary of Xerox) puts more emphasis nowadays on technologies such as predictive analytics, context engines and "Behavioral Ware". It’s much more about the software, than the sensors.

Sponsor

According to Begole, "the early Ubicomp [in the late 80’s and 90’s] was all about making machines aware of the environment." However this was "really hard because the sensors didn’t exist – or they weren’t cheap enough – and we didn’t know exactly how to use sensors to detect the environment."

But the world of sensors is "not hard anymore," Begole claimed. He pointed to accelerometers and gyros on the iPhone and Android, which "any developer can access."

Begole explained that PARC is trying to take sensor technology to the next level, by adding a layer of meaning atop the sensor data. "It’s not just the state of the world," he told me, "but what does that state mean to the user? What’s important, what’s significant about the situation? That’s what we’re trying to focus on."

He noted that new types of sensors are being developed too. For example, his Ubicomp group at PARC have been experimenting with biochemical sensors. "We haven’t done much with biochemical sensors," he added, "although we do some chemical sensing in one of the hardware labs here."

He also mentioned disposable sensors, using print, that can be put on your body like a band aid.

Here is a presentation by Bo Begole that explains PARC’s Ubicomp software developments further:

The software side of the Internet of Things is where much of the innovation is happening. PARC, HP, IBM, Cisco and others are all working hard to develop algorithms that will process and make sense of the explosion of data coming from sensors.

Discuss