San Jose Police Use Fitbit Data to Solve Murder

A new report by Wired dives deep into the case of a 90-year-old alleged murderer from California that was captured based on data from his victim’s Fitbit fitness tracking device. Tony Aiello died last week still awaiting his murder trial.

According to a report by Wired, the San Jose Police Department was able to track down 90-year-old alleged murderer Tony Aiello after it reviewed data from the victim’s Fitbit device. Aiello was arrested in late 2018 after officers reviewed data from the victim’s Fitbit, which reported a drastic rise in heart rate followed by a drop to zero. Officers were able to use the time data from the Fitbit to link Aiello to the crime.

That timeline would later become crucial. On September 19, Fitbit’s director of brand protection, Jeff Bonham, stopped by San Jose police headquarters to collect Karen’s smartwatch. He reported back to police the next day that Karen’s device had been syncing via Bluetooth every 15 minutes with its paired device. An initial analysis showed that she didn’t take any steps after 3:13 pm. The data recorded her heart rate accelerating around 3:20 pm, then taking a “precipitous drop” and ceasing altogether by 3:28. The device did not capture any more motion activity until Karen’s body was taken to the morgue. The prosecution would later allege that she was “brutally murdered in her own home while eating her last meal.” Investigators had found pieces of pizza at her feet.

Attorneys for Aiello argue that Fitbit data should not be admitted as evidence in the murder trial. They cite evidence that suggests that the Fitbit data was not reliable. They even claim that the data used to link Aiello to the crome was given a “confidence score” of zero by Fitbit’s internal algorithms.

Tony’s defense lawyers signaled that they would attack the reliability of the Fitbit data. They assembled a grab bag of disqualifications: They said Karen wore the device for only two weeks or less, and it hadn’t yet normalized to her signal; they said that Fitbit, which assigns a confidence score of 0 to 3 to its data collection, at times assigns zero confidence to the data on Karen’s device on the day the prosecution says she was murdered; and Edward Caden, one of the defense attorneys, said that what the prosecution calls a “spike” in Karen’s heart rate is more like “a pimple.” Caden even asserted that there were moments after 3:28 pm when Karen’s Fitbit seems to still report heartbeat data.

However, this isn’t the first murder case where Fitbit data has been used as evidence. In 2016, a Wisconsin judge admitted step-counting data from a murder victim. However, the judge rejected sleep data from the Fitbit, citing evidence that such data is often slightly inaccurate.

So far, just a few judges have issued rulings that discuss in detail how to handle evidence from fitness trackers. In a 2016 Wisconsin case, Fitbit data was used to eliminate the possibility that a woman was murdered by her live-in boyfriend. The judge ruled that an affidavit from Fitbit established the device’s authenticity, and allowed lawyers to introduce its step-counting data; at trial, a sheriff’s department analyst vouched for the reliability of the man’s particular device. However, the judge barred the Fitbit’s sleep data, citing a class-action suit that claims the sleep tracking could be off by as much as 45 minutes. (The murder case is now being appealed.)

Aiello died on September 11, 2019, still awaiting his murder trial.