Smart Storage: Improving the User Experience on Mobile Devices by More Intelligently Managing Data

Principal Investigator: Randal Burns, the John Hopkins University (USA) Whiting School of Engineering

Date: November 18, 2010


This study includes the COLLECTION OF PERSONALLY IDENTIFYING INFORMATION, including the names of files on your mobile computer, and web sites that you visit. Before installing the software, you should carefully read this document to understand what information is collected and the potential risks. IF YOU DO NOT COMPLETELY UNDERSTAND THE RISKS, which are explained in detail below, DO NOT INSTALL THE SOFTWARE.


The purpose of this user study is to understand what types of data people use on mobile computers, such as smart phones and laptops, how they use that data, and when network connectivity and power are available. The overarching goal of the research project is to improve the user experience on mobile computers by better automating the management of data--automatically downloading data the user is likely to use, and deleting (or suggesting for deletion) data that the user is likely to no longer need when free storage space becomes scarce.


Anyone 18 years of age or older may participate in the study. There are no further restrictions on who can fill out the questionnaire. Anyone may install the logging software who has a compatible device and operating system or who borrows a device from us. We only have a limited number of devices and will select recipients based on need and responses to the questionnaire.

We expect to have hundreds of participants who fill out the questionnaire and/or install the software.


This experiment consists of two optional parts.

The first part is a questionnaire, which asks you about how you use computers and the Internet and some specific technologies including RSS feeds and PodCasts. The questions do not request data that can be used to identify you individually. We estimate that filling out the questionnaire will take approximately 20 minutes.

In the second part, you run a logging program on one or more computers for approximately one year. Installing the program will take approximately 10 minutes. After installing the program, you will not need to interact with it again.

(If you don't have a smart phone and would like to borrow one, we have a limited number available. In this case, you need to fill out the questionnaire and indicate that you would like to borrow one. Based on the responses, we will select a number of people.)

The logging software collects data about the files you use, the programs you run, your network connectivity and battery status. Specifically, the collected data are:

  1. File accesses (the filename of the accessed file, the type of access, create, delete, read, write, and the time of the access; the contents of the file will not be recorded);

  2. Programs run (their names, the time at which they were started, and the time at which they exited);

  3. Battery status (the available power as reported by the operating system and the time the sample was taken, when the device is attached to a power source and when it is detached);

  4. Network connection statistics (the name of the network, anonymized using a one-way hash (explained in the section CONFIDENTIALITY, below), e.g., the cell phone provider's common name or the WiFi access point's SSID, when a connection is established, when a connection is disconnected, the number of bytes transferred, the connection's signal strength, and the time of the sample);

  5. Available networks (the identity of networks, anonymized using a one way hash, available in the vicinity of the device, but which the device has not connected to, and the time stamp at which the sample was made);

  6. Connected cell tower (time at which the device connect to the cell tower and the tower's identification including its location area code, the cell identifier, the network code and the country identifier, anonymized using a one-way hash, except the country identifier);

  7. User activity (when the user starts interacting with the system, and when the user stops interacting with the system, e.g., when the screen saver is deactivated or activated);

  8. System boot, shutdown, suspend and resume (the time when the system boots, when the time when the system is turned off, when the system is suspended and when it is resumed);

  9. Web sites that the user visits, anonymized using a one-way hash for each URL component except for the name of the file, e.g., would be reported as something along the lines of http://a1040356.3275c2/0016fff24/pic1.jpg; and,

  10. Warnings and error messages emitted by the logging program (the time at which the message was emitted and the message itself; for debugging purposes).

The program is designed to be lightweight so as to minimally interfere with you. Moreover, the program will be run with a low priority.

Approximately once per day, the logging program will upload the data over an encrypted connection to, a server under PI Burn's control. This will only be done if:

  1. There is an ethernet or WiFi connection available; a cellular connection will never be used to avoid the chance of the logging program incurring monetary charges.
  2. The user appears to not be interacting with the system (e.g., the screen saver is active).
  3. The server's reported identity matches the expected identity (this is done using cryptographic techniques and ensures that the data is not accidentally reveal data to the wrong server).

If the data is successfully uploaded, the synchronized data will be deleted from the device thereby reducing the storage space used by the logging program on the device and avoiding the case that someone who later obtains the device gets access to the logged data.

The data will be identified by a random, secret identifier that is generated on the device when the program is first run.

When the experiment is completed, an updated version of the program will be made available, which should automatically be made available to you using the operating system's usual software update mechanism. This version of the program will delete any remaining logged data, disable logging, and remove the logging software.


The risks associated with participation in this study are minor; we try to be aware of, and carefully avoid, any problems that could arise for participants as a result of taking part in this research. Disclosure of data collected about you (detailed above in the section PROCEDURES) is the primary risk. The data that may have personally identifying information includes:

  1. File names (the standard email client on the Nokia N900 includes the email address of the email account in the file name when saving email messages);
  2. URLs of web pages the user visited; and,
  3. Location information (names of WiFi access points and cellular towers).

If collected data is exposed, you may be embarassed: the data may include URLs of web sites that embarass you, e.g., pornography.

A person who accesses your device (e.g., a thief) is able to access recently logged data. As data is anonymized prior to it being saved, the amount of information an attacker could gain is reduced. Further, this is limited to data logged since the last upload, which typically occurs every 24 hours, since logged data is deleted from the device. Some information an attacker could learn include: The names of files that you have recently deleted. If the person were interested in knowing whether you were at a particular location and knew the identity of cell phone towers in that area, that person could verify whether you were there at a particular time.


Any study records that identify you will be kept confidential to the extent possible by law. The records from your participation may be reviewed by people responsible for making sure that research is done properly, including members of the Homewood Institutional Review Board of the Johns Hopkins University, or officials from government agencies such as the National Institutes of Health or the Office for Human Research Protections. (All of these people are required to keep your identity confidential.) Otherwise, records that identify you will be available only to people working on the study, unless you give permission for other people to see the records.

To protect your privacy, we partially anonymize the URLs and location information using a one-way hash (as described in the section PROCEDURES). A one-way hash is a deterministic procedure for transforming data. It has two important properties: given the hashed data, it is impractical to recover the original data; and, the same data always results in the same transformation. These properties allow us to determine how often participants frequency a particular location or web site without being able to identify the web site or location. However, because the hash transforms the same data in the same way, it is possible to verify the existence of particular data. For instance, determining what corresponds to "2q34lkjgfd.9808" is difficult. However, if someone with access to the data wants to determine whether a trace includes particular data, e.g., "," it is only necessary to compute the hash corresponding to "," perhaps, "2q34lkjgfd.9808," and checking if that data appears in the trace.

Data collected from you will be stored on password protected computers. The data collected by the logging program will be transferred using an encrypted connection. The identity of the server will also be cryptographically verified. Ten years after the study has been completed, the data will be deleted.


You will not receive any direct benefits from participating in this study. This study may benefit society if the results lead to a better understanding of how data is used on mobile devices. Specifically, this study tries to increase data availability, improve the performance of data access, improve battery life, decrease internet connectivity costs, and facilitate managing data by determining which files to prefetch and which files are least valuable.


Your participation in this study is entirely voluntary: you choose whether to participate. If you decide not to participate, there are no penalties, and you will not lose any benefits to which you would otherwise be entitled.

If you choose to participate in the study, you can stop your participation at any time, without any penalty or loss of benefits. If you want to withdraw from the study, you need either stop filling out the questionnaire, or uninstall the software on your devices.


You will not receive any payment or other compensation for participating in this study.


You can ask questions about this research study now or at any time during the study, by talking to the researcher(s) working with you, or by emailing Neal Walfield (, the student investigator for this study. If you have questions about your rights as a research subject, or feel that you have not been treated fairly, please call the Homewood IRB at Johns Hopkins University at (410) 516-6580.