Undistracted Driving:A Mobile Phone that Doesn’t DistractJanne LindqvistCarnegie Mellon University5000 Forbes AvenuePittsburgh, PA 15213 United [email protected] HongCarnegie Mellon University5000 Forbes AvenuePittsburgh, PA 15213 United [email protected] driving is a major problem that leads to unnec-essary accidents and human casualties everywhere in theworld. The ubiquity of mobile phones is one cause of dis-tracted driving. In United States alone, operating mobilephones while driving has been cited as a factor in crashesthat have led to 995 deaths and 24,000 injuries in 2009. Tomitigate the problem of distracted driving caused by mo-bile phones, we propose using context-awareness to imple-ment burden-shifting, time-shifting, and activity-based shar-ing. Although the first two concepts have been introducedbefore in the research literature and the latter two are novel,none of these concepts have yet been explored in the contextof mobile phones and driving. We present our initial inter-action designs for these concepts on the Android platform.Categories and Subject DescriptorsH.5m [Information Interfaces and Presentation (e.g.,HCI)]: Miscellaneous; K.4.0 [Computers and Society]:GeneralGeneral TermsDesign, Human FactorsKeywordsdistracted driving, mobile phones, human interruptibility,context-aware communications1. INTRODUCTIONToday, a mobile phone user is reachable virtually any timeand any place. While there are many benefits in being con-nected all the time, it can also lead to interruptions at un-desirable times. In some cases, these interruptions are in-convenient, such as when in a meeting, but in other casesthey can be dangerous, such as when driving. For example,during a typical daylight moment in the US in 2009, 9%Permission to make digital or hard copies of all or part of this work forpersonal or classroom use is granted without fee provided that copies arenot made or distributed for profit or commercial advantage and that copiesbear this notice and the full citation on the first page. To copy otherwise, torepublish, to post on servers or to redistribute to lists, requires prior specificpermission and/or a fee.HotMobile’11 March 1-2, 2011, Phoenix, Arizona, USA.Copyright 2011 ACM 978-1-4503-0649-2/XX/XX ...$10.00.of all drivers were using a hand-held or hands-free phonewhile driving [29]. Using mobile phones while driving in theUnited States has been indicated [28] as a factor in crashesthat have led to 995 deaths (in 867 crashes, 18% of all fa-tal distracted-driving crashes) and 24,000 injuries (5% of allpeople injured in crashes) during the year 2009. It is not sur-prising that a body of research [22, 4, 3, 28, 29] has shownhow using mobile phones distracts the driver from their pri-mary task of driving. Due to this extensive research, manystates [12] in the US have banned texting or using hand-held phones while driving, while the federal government hasbanned truck drivers from texting [27].We observe that people need and want to use their mo-bile phones even while driving [25], for example, to coor-dinate meetings or communicate directions. Towards thisend, mobile phone and car manufactures have designed var-ious hands-free systems utilizing e.g. Bluetooth and voice-control. Unfortunately, research has indicated [4] that evenif the mobile phone is operated hands-free, it can neverthe-less distract the driver and may therefore cause crashes. Wealso note that with today’s mobile phones, the caller doesnot know what the person who they are trying to call is do-ing, for example, are they driving or not. On the other hand,sometimes it is important to be interrupted, for example ifthere is something urgent related to work or during a familyemergency. Our observation is that some of these interrup-tions introduced by a mobile phone while driving could bedeferred or resolved in other ways.To address the problems outlined above, we propose fourtechniques that reduce the need to operate a mobile phonewhile driving. Our approach is to leverage context-awareness,such as location and movement of the call recipient, andthe identity of the caller to implement burden-shifting, time-shifting, and activity-based sharing. This information can beused to 1) give callers appropriate information about recip-ients’ context without recipients actually needing to answerthe phone, 2) let callers communicate with recipients whenthe situation is appropriate, 3) respond automatically whenpossible, 4) defer messages and send pre-planned messages,and 5) share estimated arrival time. In designing these ap-proaches, we need to balance multiple conflicting require-ments: the new system should be easy to use and configure,establish more efficient communication, substantially reduceundesired interruptions, be privacy sensitive, and interruptthe driver when the situation demands it. In the rest of thispaper, we further motivate our approach, present our designand contrast it to related work, and discuss engineering andimplementation issues and limitations.2. MOTIVATIONTo further motivate our approach, we discuss a commonscenario that can involve driving and using a mobile phone:picking up a friend with your car. With ubiquity of mobilephones, it is common that people are flexible when schedul-ing meetings and such [16]. For example, without our pro-posed system, we can imagine people agreeing “I’ll call whenI’m near your apartment so you’ll know when to come out.”Arranging for a pickup can involve the following steps 1) ne-gotiating the (approximate) time, 2) informing the personto be picked up that the driver is leaving or on the way, and3) informing that that person that the driver is nearby orat the pickup location. Alternatively to 3), sometimes theperson waiting for the pickup might get anxious where thedriver is and 4) call the driver, and the driver might be stuckin the traffic and 5) call to inform of late arrival.We use this scenario to illustrate the key concepts we uti-lize in this paper: burden-shifting (BU), context-awareness(CA), time-shifting (TS) and activity-based sharing (AB).Bob calls Alice whether she could pick him up when sheleaves work. Alice agrees with Bob that she’ll pick himup around 6 PM. After the call has finished, Alice is pre-sented with an after-call interface where she chooses theoption “schedule a pickup” (TS). She also chooses to shareher estimated-time-to-arrive (ETA) with Bob near the timeof the