On Tuesday, Samsung announced that it would begin mass producing its first-ever "Bio-Processor." This is a new processor specifically designed for the monitoring and computing of physiological metrics such as heart rate, skin temperature, body fat, heart rhythm, muscle mass, and stress levels. In short, the human body will spew data like never before.
The Bio-Processor isn't quite a processor as we normally think about it, which is a generic piece of hardware (a CPU) that handles basic arithmetic, logic, and input-output operations based on instructions delivered via a computer program that may or may not involve more advanced I/O hardware, memory devices, GPUs, and-or any of the other things that add up to being a "complete" computer. It's more properly a "system on a chip" (SOC), which is much as it sounds: the processor plus all of the other junk all loaded onto a single discrete hardware unit.
The actual CPU of the Bio-Processor is an ARM Cortex-M4. This is a pretty common processor for embedded microcontrollers, particularly for industrial and scientific applications, and is very cheap. It so happens I have one in front of me stuck to a Texas Instruments development board (sort of like a beefed-up, data-able Arduino) and I think the whole thing cost around $12 with shipping.
That's the thing about the M4. It's meant for data and data is what health monitoring is all about. The Bio-Processor puts a bunch of sensors on the same chip as the M4 itself along with the system's memory units, security units, and an analog front-end (so it can read analog data and handle it digitally). The Samsung Bio-Processor product page advertises as well the inclusion of a DSP (digital signal processing) unit, which is a necessary feature for data applications and also happens to come integrated with ARM's M4 processor anyway.
So, now we have all of this data. Cool. What now?
That's the implicit question in the whole health-monitoring rush. The answer is implied too, of course: "be healthier." Assuming that you are not currently in an ICU, is knowing the volume of your organs in real-time (via photoplethysmogra) or your galvanic skin response (a super-problematic measurement of the skin's varying electrical conductance properties) or even your ECG actually useful information? Or is it just another thing to obsess about?
As you answer, it's worth considering the enormous bets being placed on health-monitoring technology by Samsung, Apple, and many others. Consumers obsessing over health data is the new thing, an inescapable psueudoscience supernova, which is maybe not a great indicator of the state of new things circa 2016.
Des Spence, a UK general practitioner, put this well in a piece last April in the BMJ, writing that "health and fitness have become the new social currency, spawning a 'worried well' generation."
Most medical research and diagnoses are based on isolated readings taken in medical clinics in symptomatic, older, high risk individuals, by doctors who can interpret results—not by young, asymptomatic, middle class neurotics continuously monitoring their vital signs while they sleep. So what will users of these apps discover? How common brief arrhythmias are in the normal population? How often our blood pressure might be high? How widely normal oxygen saturations can vary? The variation in the heart rate of an intrauterine baby? What happens if these gizmos malfunction or are placed in the wrong position? How will it change our management? Who can interpret the results? What if parents want to start monitoring their children? Where’s the evidence that these things will improve diagnosis?
It's a good rant, but what he fears is probably unavoidable: "A Wild West approach to development is playing out and will use the advertising classic—fear—to sell product. War, pestilence, and famine are all out to grass; technology, medicine, and overdiagnosis are the new riders of the Apocalypse."