We’re back from Relativity Fest in Chicago and ready to kick off a new decision for you in episode 86 of Case of the Week.
This week’s decision dated September 23, 2022 is in the In re 3M Combat Arms Earplug Prods. Liab. Litig., 2022 WL 4448917 (N.D. Fla. 2022) matter. We’ll discuss whether a party can be compelled to disclose noise exposure data automatically captured by smartphones and smartwatches. The decision comes from United States Magistrate Judge Gary R. Jones.
Keep reading or watch the video to understand the eDiscovery issues.
Good morning, and welcome to Episode 86 of our Case of the Week series, published in partnership with ACEDS. My name is Kelly Twigger. I am the Founder and CEO of eDiscovery Assistant, as well as the Principal at ESI Attorneys, and I’m very happy to be with you today. Thanks so much for joining me.
It was so great to see so many of you at Relativity Fest last week in Chicago. Really appreciate how much support we received for our Case of the Week broadcast, and a very special shoutout to Monica Schultz at Nyemaster Goode who’s a big fan of the broadcast. Thanks so much for being here with me each week. I really appreciate it.
All right let’s jump into this week’s decision from In re 3M Combat Arms Earplugs Product Liability Litigation. That is a mouthful. This is the 16th discovery decision in this case from Magistrate Judge Gary Jones, who is the United States Magistrate Judge in the Northern District of Florida. Judge Jones is a colleague of mine, a fellow member of the Planning Committee at the University of Florida Law eDiscovery Conference that’ll be held in February. He is a very thoughtful and prolific writer on discovery issues and has 39 decisions in our eDiscovery Assistant database.
In this particular decision, on a motion to compel the production of noise data, Judge Jones denied the motion from the defendants, finding that the defendants failed to establish that the noise exposure data would be important in resolving the issues in the case, and that the burden and expense of collecting and reviewing the hearing data outweighed the likely benefit to production. So, the motion failed on proportionality grounds.
All right, today’s decision from Judge Jones is dated September 23rd, 2022. As always, we assign issues to each decision in our case law database. Our issues this week include internet of things, self-collection, wearables, audio, mobile device, failure to produce, and proportionality.
What are the facts that we’re talking about here? Well, I mentioned already that we’re before the Court on defendant’s motion to compel the plaintiffs to produce noise exposure data that was automatically recorded by the operating systems of plaintiff smartphones and smartwatches. The defendants contended that the noise exposure data was relevant to the plaintiff’s hearing injury claims and could be collected with a minimal burden, and they requested it formally in an RFP on March 3rd of 2022.
Now, this request is specifically directed to the second wave of plaintiffs, or Wave 2 as the Court refers to it, and was not propounded in overall discovery or directed to Wave 1 of the plaintiffs. So that’s important as to the timeliness issue.
The defendants argue that the data is relevant and proportional because the data will show how loud and how long the plaintiffs used their headphones, how loud and how long the plaintiffs were exposed to environmental noise, and how frequently the plaintiffs were notified by their phones that they were being exposed to dangerous levels of noise. Now, as I mentioned, the defendants did not request this information during bellwether discovery or during discovery to the Wave 1 plaintiffs.
Now, to assist with the production, when the defendants made their request, they provided plaintiffs with written instructions on how to locate the information that is stored by the operating system for Apple smart devices, which is referred to as iOS, as well as how to capture that information by taking a screenshot. That’s something that’s unusual. We don’t see that a lot with requests for production — a party providing specific steps on how to capture data. But that’s important here, where the defendants believe that there can be a minimal burden for the plaintiffs to self-collect their own data and provide it to counsel for production.
Now, the data specifically sought by the request is the noise exposure and hearing data, and related information, that is recorded on Apple iOS devices. It’s only recorded on Apple iOS devices that are using an operating system post 2019.
The Court notes that on devices where Apple Health data is enabled, the devices have the capability to record environmental sound levels, which Apple describes as “sounds in your environment, measured at a weighted decibels” and then recorded when the noise app on an Apple Watch is synced to the iOS device.
You got a few things that have to happen here factually in order for this information to even exist before we start determining relevance and proportionality:
- The user has to have an Apple iOS device.
- They have to be using the operating system post 2019 that even had the capabilities with the Apple Health app to sync and be able to capture this environmental noise, as well as the warnings that supposedly the phone emits if the user is exposed to high decibels of noise.
- Then you’ve got to have all of this enabled and synced, especially if the user is also using an Apple Smartwatch. The Apple Smartwatch has to be synced to the phone in order for that information to be collected from that particular device.
Okay, so how do the plaintiffs respond? Well, one plaintiff from Wave 2 produced the information, but the remaining plaintiffs objected to proportionality. Some of those plaintiffs that objected stated that they had conducted an investigation and had no responsive data. However, at depositions, some of the individuals, those same plaintiffs, testified that they had not looked for the data that had been responded to by counsel.
The defendants here are essentially asking for a court order for all of the plaintiffs to conduct an investigation for the noise data. Both parties submitted affidavits from computer forensic experts detailing how information is captured, what information is captured, and the complexity with retrieving that information, as well as some difficulties in determining whether the information could in fact be relevant or potentially admissible in the case.
What’s the Court’s analysis on that set of facts? The Court outright dismisses the plaintiff’s objection to the timing on the motion, and notes that while the timing is important here, i.e., that this request was not made in bellwether discovery or to the Wave 1 plaintiffs, that there is not a sufficient reason alone for denying the motion here. So that is off the table.
The Court then next turns to the proportionality analysis and reviews the six proportionality requirements of Rule 26(b)(1). Those include the importance of the issues at stake in the action, the amount of controversy, the party’s relative access to relevant information, the party’s resources, the importance of the discovery in resolving the issues, and whether the burden or expense of the proposed discovery outweighs its likely benefit.
What the Court ultimately finds is that the defendants failed to establish that the noise exposure data would be important to resolving the issues in this case, and that the burden and expense of collecting and reviewing the hearing data outweighs the likely benefit to producing the data. As such, the motion fails on proportionality grounds.
What is the Court’s reasoning regarding that analysis? Well, they gave several reasons, all of which are very factually based on the availability of the information as well as the cost to be able to retrieve that information.
The Court starts with the fact that it did not feel that the defendants had shown that all or any of the Wave 2 plaintiffs actually possessed an Apple device that was capable of storing the information they seek. We talked about the fact that both of the expert affidavits indicated that this noise information is only captured by Apple iOS devices and only captured by those Apple devices that had operating systems post 2019 when the Apple Health data app was created and released.
The Court’s analysis following the review of the expert affidavits is that only those Apple devices with that post-2019 iOS operating system would have collected the information sought, and that the defendants failed to issue an interrogatory requesting what devices each plaintiff used and whether they had a device that might have even collected the requested data.
The Court also found that defendants had not shown that the data was available from any other personal electronic device, including an Android device. By virtue of the fact that the defendants didn’t even know which ones of the plaintiff had collected this information, it was not appropriate for them to have the Court order them to search for it.
The second issue that the Court raises is a privacy issue in the data on the associated devices. The Court found that even if a Wave 2 plaintiff possessed a post-2019 iOS device, that collection and review of the data with industry-standard tools would result in the production of other private health data from the user and potentially of non-parties with whom the user shares a phone or other device, such as family members.
Essentially, the Court is raising that privacy issue here, that we’ve seen many times addressed in data from a mobile device. When you collect information using Cellebrite or other court approved, I guess I’ll say, forensic processes, you’re going to collect a lot more data from the mobile device than what you actually need. The Court finds that the privacy interest from the folks with accounts on that phone is a factor in the proportionality determination here.
The Court notes here that there’s limited information available about the user of the device at the time the noise data was recorded or how the device was being used. This is another basis as to why this request is not proportional. The Court notes that it is not uncommon for others to use a device to make a phone call, for music, or for navigation in the car. In fact, just this last weekend, I was driving my daughter to Terror in the Corn at one of our local farms, and she had my phone for the entire trip, both playing music as well as navigating us to the particular location out in the middle of nowhere. The Court makes a specific determination here of actual usage of how people use their phones — which, in my case, would have indicated if my daughter turned the music up to a certain decibel range that might have triggered a warning on my phone, that would have been her conducting that, and not me.
If I was the plaintiff in this litigation and my daughter had created that information, it would not be relevant here. The Court notes that it’s really difficult to discern who was using the device at the time that any kind of noise information that might be relevant would have been created.
In addition to the fact that the hearing data doesn’t disclose who was using the phone, the data also does not provide any information about where the device was located at the time the hearing data registered, and it doesn’t provide any information on where the device was located in relation to the user, i.e., who was holding the device? Where was it sitting? Was my daughter holding it in her hand? Was it sitting on the dashboard? Was it in the backseat coming from someone else? All of that would impact or have relevance to the impact of that noise data on my hearing, again, if I’m the plaintiff.
The Court notes in furtherance of the proportionality analysis that the reliability and usefulness of the hearing data is further diluted because the requested data does not disclose whether the user was using headphones, a Bluetooth speaker, an automobile speaker through use of CarPlay, or any other kind of connected device.
That’s important because the difference between hearing data registered when using Apple earbuds and hearing data registered while the device was connected to speakers is vastly different in terms of the effect on an individual plaintiff’s hearing. There’s no ability to determine that data from the phone. Whatever hearing data exists on the phone, there’s really no ability to determine whether someone was listening to that directly into their ears on headphones or whether it was playing 20 feet away from them.
The Court notes, “It is highly doubtful that any hearing expert would associate hearing loss with sound levels from a speaker where the user is not in close proximity to the connected device. The data from the self-collection defendants have identified would provide no information as to whether the user even viewed the hearing health information or took immediate action to turn down the volume in response to a headphone or noise alert level.”
All right, next basis from the Court is that the noise alert data that the defendants seek does not provide any evidence of exposure to hazardous noise. Essentially, if you turn up the volume on an Apple iOS device with a post-2019 operating system, there may be or are alerts that are issued on the phone for noise exposure. But the question is, to what extent, if there are any of those alerts, what was the range of noise that was heard by the actual plaintiff? There’s no data about that.
The Court notes that the noise alerts recorded on the iOS devices track the times when the iOS device automatically reduced noise that may have exceeded 85 decibels, but the Court notes that even defendants’ expert recognizes that continuous noise is not hazardous until it exceeds an eight-hour TWA of 85 decibels. Because the data, at most, evidences exposures to 85 decibels of noise during isolated times over the course of a week, it’s not particularly useful evidence of the plaintiff’s injuries.
In effect, the Court is saying, look, in order to have a sustained injury due to noise exposure and to have this evidence be relevant, you’re going to have to show a continuous period of time of exposure at these 85 decibels or north of that. At most, what you’re going to see is short periods of time where that happened. We don’t have the evidence that we need in order for this to be relevant to plaintiff’s injuries.
The Court also then turns to potential inherent problems with the reliability of the noise data. The Court notes that the quality and accuracy of the recording of the data by the device is dependent upon the age and wear and tear of the device as well as how the device is held and positioned.
Further to that, the Court notes that devices that have a case on them, which is pretty much every device, will alter the factory-designed response of the microphone, which in turn affects how accurate the noise measurements might be. You can have 85 decibels of noise coming out of the phone itself based on the volume it’s turned up to, but a case or the positioning of the phone, or me holding it in my hand with my hand over the microphone, would all change the output that would go to a particular plaintiff.
The Court also rejected the defendant’s argument that this evidence is relevant and that the issues that the Court is raising really look at admissibility as opposed to relevance. The Court says that’s not the case here: “The fundamental problems with the data show, however, and inform the Court’s analysis of proportionality, because the collection, review, and production of data that has minimal to no benefit in resolving the issues in these cases certainly fails any measure of proportionality.”
The Court also notes that the burden and expense of producing the data far outweighs any benefit. The defendant’s proposed self-collection protocol assumes that each plaintiff would be able to follow it without expert assistance, which the Court doubted. If each plaintiff cannot follow those instructions, that leads each plaintiff back to a full forensic collection of the device, which of course can be done remotely, but still leads to an additional burden of time and expense that the Court finds is not warranted given the limited value of the data.
The Court really ends with this quote, which I think is pertinent to its analysis from a factual perspective and as we think about this limited data that’s available from a mobile device:
The Court concludes that the data defendants seek at best reflects data about the device itself and not data about how the plaintiffs were affected by noise generated by the device. Because the data will not show who was using the device, what connected external device was being used, or even where the user was located in relation to the device when a hearing event was registered, collection, review, and production of the data is not proportional, and therefore, defendant’s motion to compel is denied.
Takeaways from our case today. This decision from Judge Jones really considers how people use their electronic devices and the legitimacy of the data that defendants sought to collect as part of the proportionality analysis. We’ve seen other decisions where data from wearable devices, data from mobile devices, has been deemed to be relevant, and the considerations of admissibility will be considered further down the line. This is one of the first opportunities that we’ve seen of a court doing an analysis of the viability of data to be collected and its actual use from a proportionality consideration under Rule 26.
The data that’s sought here is a very specific subset of information that’s available on a mobile device. It’s even more specific in that it is only available on specific Apple mobile devices using that post-2019 iOS operating system.
Now, I think we saw from the facts here that essentially defendants didn’t properly set up the staging to get this data. It would have been more appropriate for them to determine the group of plaintiffs that had the data first, and presumably to address some of the issues that the Court raised in its determination on proportionality.
Now, I didn’t review the submissions from the experts on this decision in prep for Case of the Week today, but I think that there was likely more information in the plaintiff’s expert declaration that raised a number of these issues for the Court to consider on the actual viability of the data and how relevant it was going to be to the plaintiff’s injuries. That is really what tipped the balance of the Court’s analysis here, is that, frankly, the data was just too circumspect to really be able to create any relationship between this noise data on the phone and what the plaintiffs actually heard relative to the injuries created.
Now, what this case does give you is an excellent roadmap for how to lay out information and expert affidavits to provide the Court the knowledge of how data is recorded. I didn’t go over all of the facts for you today, but the Court goes through a very in-depth discussion of what data is recorded, how it’s recorded, what the decibel readings need to be — a lot of factual information that clearly was provided by expert affidavits here for Judge Jones.
This case definitely shows that the parties — where there’s normally a cursory relevance determination that can be made to allow for the production of data — a party can argue proportionality with specific facts, as the plaintiffs did here, to be able to combat that relevance determination and still preclude production.
Take this case as one that allows you to … when you’re looking at very specific evidence from ESI, you want to be able to argue that you’ve got a full set of data that does not have reliability considerations in order to be able to get it under a relevance determination. Otherwise, a proportionality analysis by the other side that can poke holes in the reliability of that evidence against the cost and expense, cost and time required, or the burden to produce it, is what’s key to the proportionality analysis.
Okay, that’s our Case of the Week for this week. Thank you so much for joining me. We’re back again next week with another decision from our eDiscovery Assistant database.
If you’re interested in doing a free trial of our software, sign up to get started, or you can email us at firstname.lastname@example.org, and we’ll work to get you set up.
Thanks so much. Have a great week.