Opinion by Judge CALLAHAN;1 Separate Opinion by Judge SILVERMAN; Partial Concurrence and Partial Dissent by Judge HALL: Dissent by Judge CALLAHAN.
ORDER AND OPINION
ORDER
The petition for panel rehearing is granted in part and denied in part. The Opinion filed August 21, 2009, and appearing at 578 F.3d 1084 (9th Cir.2009), is withdrawn. It may not be cited as precedent by or to this court or any district court of the Ninth Circuit.
The superseding opinion will be filed concurrently with this order.
OPINION
CALLAHAN, Circuit Judge:
Mike Rutti sought to bring a class action on behalf of all technicians employed by Lojack, Inc. (“Lojack”) to install alarms in customers’ cars. He sought compensation for the time they spent commuting to worksites in Lojack’s vehicles and for time spent on preliminary and postliminary2 ac[1049]*1049tivities performed at their homes. The district court granted Lojack summary judgment, holding that Rutti’s commute was not compensable as a matter of law and that the preliminary and postliminary activities were not compensable because they either were not integral to Rutti’s principal activities or consumed a de minimis amount of time. We affirm the district court’s denial of compensation under federal law for Rutti’s commute and for his preliminary activities. However, we vacate the district court’s grant of summary judgment on Rutti’s claim for compensation of his commute under California law and on his postliminary activity of required daily portable data transmissions, and remand the matter to the district court for further proceedings consistent with this opinion.
I.
A. Facts
Rutti was employed by Lojack as one of its over 450 nationwide technicians who install and repair vehicle recovery systems in vehicles. Most, if not all of the installations and repairs are done at the clients’ locations. Rutti was employed to install and repair vehicle recovery systems in Orange County, and required to travel to the job sites in a company-owned vehicle. Rutti was paid by Lojack on an hourly basis for the time period beginning when he arrived at his first job location and ending when he completed his final job installation of the day.
In addition to the time spent commuting, Rutti sought compensation for certain “off-the-clock” activities he performed before he left for the first job in the morning and after he returned home following the completion of the last job. Rutti asserted that Lojack required technicians to be “on call” from 8:00 a.m. until 6:00 p.m. Monday through Friday, and from 8:00 a.m. until 5:00 p.m. on Saturdays. During this time, the technicians were required to keep their mobile phones on and answer requests from dispatch to perform additional jobs, but they were permitted to decline the jobs.3 Rutti also alleged that he spent time in the morning receiving assignments for the day, mapping his routes to the assignments, and prioritizing the jobs. This included time spent logging on to a hand-held computer device provided by Lojack that informed him of his jobs for the day.4 In addition, it appears that Rut-ti may have completed some minimal paperwork at home before he left for his first job.
During the day, Rutti recorded information about the installations he performed on a portable data terminal (“PDT”) provided by Lojack. After he returned home in the evening, Rutti was required to upload data about his work to the company. This involved connecting the PDT to a modem, scrolling down a menu on the PDT [1050]*1050until he encountered an option labeled “transmit,” and selecting this option to initiate the upload process. The transmissions had to be done at home because it required the use of the modem provided by Lojack. Rutti was required to make sure that the transmission was successful, and there is evidence in the record that it often took more than one attempt to successfully complete a transmission. Lo-jack’s Installer Training Manual instructed technicians not to transmit their PDT data ten minutes before or after the hour because the corporate computer system is automatically reset at those times. The Manual further instructed technicians to wait an hour if they have technical difficulties and that after two unsuccessful attempts they should call the host computer and document the date, time, PDT error message, number called from, and any specific error message, dial tone, or busy signal heard over the phone line.
B. Procedural History
On April 5, 2006, Rutti filed this putative class action on behalf of himself and similarly-situated technicians asserting that under the Fair Labor Standards Act, 29 U.S.C. §§ 201-19 (“FLSA”), and under California law, Lojack had unlawfully failed to compensate for commuting and “off-the-clock” work. After the parties had engaged in considerable discovery, Lo-jack moved for partial summary judgment and Rutti sought class certification. The district court decided to rule on the motion for partial summary judgment before addressing class certification, citing Wright v. Schock, 742 F.2d 541, 544 (9th Cir.1984) (“It is reasonable to consider a Rule 56 motion first when early resolution of a motion for summary judgment seems likely to protect both the parties and the court from needless and costly further litigation.”). On August 16, 2007, the district court issued its order granting in part and denying in part Lojack’s motion for partial summary judgment. The order disposed of all federal claims and denied Lojack’s state law claim for compensation for commuting. The district court subsequently issued an order dismissing the remaining state law claims for lack of subject matter jurisdiction. Rutti filed a timely notice of appeal.
II.
Rutti’s appeal is from a grant of summary judgment and accordingly, we “must determine, viewing the evidence in the light most favorable to ... the non-moving party, whether there are any genuine issues of material fact and whether the district court correctly applied the substantive law.” Olsen v. Idaho State Bd. of Medicine, 363 F.3d 916, 922 (9th Cir.2004). “We may affirm on any ground that is supported by the record.” Id.; ACLU of Nevada v. City of Las Vegas, 466 F.3d 784, 790 (9th Cir.2006).
III.
Rutti’s appeal raises three major issues: (1) whether Rutti’s commute in a Lojack vehicle was compensable under federal or state law; (2) whether Rutti’s off-the-clock activities were either not part of his principal activities for Lojack or were de minim-is, and thus not compensable; and (3) whether under the “continuous workday” doctrine Rutti’s workday started at his home in the morning before he commuted to the first job and extended to his return home.5 We agree with the district court’s [1051]*1051treatment of all of these issues except as to its grant of summary judgment on Rut-ti’s state cause of action for his commute time and his mandatory off-the-eloek PDT transmissions.6
A. Rutti is not entitled to reimbursement under federal law for commuting
Rutti offers two arguments in support of his claim that he is entitled to compensation for commuting in the vehicle provided by Lojack. First, he asserts, based on a United States Department of Labor letter dated April 3, 1995, that he is entitled to compensation because his use of Lojack’s vehicle to commute was not voluntary, and amounted to a condition of his employment. Rutti’s second argument is that the restrictions Lojack placed on his use of the vehicle rendered the commute compensable. We do not find Rutti’s arguments persuasive.
1. Pursuant to the Employment Commuter Flexibility Act, use of an employer’s vehicle to commute is not compensable even if it is a condition of employment
Rutti’s first argument is that because he is required to commute in the vehicle provided by Lojack, he did not voluntarily agree to the arrangement and is therefore entitled to compensation. This argument is based on a Department of Labor letter dated April 3, 1995, which states that an employee need not be compensated for the time spent commuting when “driving the employer’s vehicle between the employee’s home and customers’ work sites at the beginning and ending of the workday is strictly voluntary and not a condition of employment.” U.S. Dep’t Lab. Op. Ltr. (April 3, 1995). Rutti reads this letter as holding that when the use of employer’s vehicle to commute is not “strictly voluntary” and is a “condition of employment,” then the employee must be compensated for the commute time.
Even assuming that Rutti might have been entitled to compensation for his commute under the April 3, 1995 letter at the time it was written, his claim to compensation does not survive the passage of the Employee Commuter Flexibility Act (“ECFA”), 29 U.S.C. § 254(a)(2). The language of the ECFA and its legislative history compel the conclusion that the requisite “agreement” concerning the use of an employer’s vehicle to commute may be part of the employee’s employment.
In 1996, Congress amended the Portal-to-Portal Act by enacting the ECFA. The statute provides that an employer need not compensate an employee for the following activities:
(1) walking, riding, or traveling to and from the actual place of performance of the principal activity or activities which such employee is employed to perform, and
(2) activities which are preliminary to or postliminary to said principal activity or activities,
which occur either prior to the time on any particular workday at which such employee commences, or subsequent to [1052]*1052the time on any particular workday at which he ceases, such principal activity or activities. For purposes of this subsection, the use of an employer’s vehicle for travel by an employee and activities performed by an employee which are incidental to the use of such vehicle for commuting shall not be considered part of the employee’s principal activities if the use of such vehicle for travel is within the normal commuting area for the employer’s business or establishment and the use of the employer’s vehicle is subject to an agreement on the part of the employer and the employee or representative of such employee.
29 U.S.C. § 254(a)(emphasis added).
The ECFA’s language states that where the use of the vehicle “is subject to an agreement on the part of the employer and the employee,” it is not part of the employee’s principal activities and thus not compensable. Id. There is no suggestion that the agreement cannot be a condition of employment. Indeed, it would appear that the most logical place to record an agreement between an employee and an employer concerning the use of an employer’s vehicle is in the employee’s employment contract. One would expect Congress to specify if it did not intend that the statute have such a likely result. Accordingly, we find the plain meaning of the statute does not support Rutti’s position. See K & N Eng’g, Inc. v. Bulat, 510 F.3d 1079, 1081 (9th Cir.2007) (“Statutory interpretation begins with the plain language of the statute. If the text of the statute is clear, this court looks no further in determining the statute’s meaning.”) (internal citations omitted); see also Adams v. United States, 471 F.3d 1321, 1323 (Fed.Cir.2006) (holding commute in government vehicle noncompensable, though condition of employment).
A review of the legislative history confirms the plain language of the ECFA. Both the sponsors and the opponents of ECFA recognized that the requisite agreement could be part of an employee’s conditions of employment. The Report on the bill (H.R.1227) stated that the bill “does not require a written agreement, this requirement may be satisfied through a formal written agreement between the employee and employer, a collective bargaining agreement between the employee’s representatives and the employer, or an understanding based on established industry or company practices.” H.R.Rep. No. 104-585, at 4 (1996). The minority report objected that the bill “permits an employer to compel an employee to agree to use the employer’s vehicle for commuting purposes, as a condition of employment,” and commented that the majority had rejected an amendment that provided that the agreement “must be knowing and voluntary, and may not be required as a condition of employment.” H.R.Rep. No. 104-585 at 8. The author of the bill responded that in some instances an employee’s use of the employer’s vehicle could be a condition of employment “depending on the agreement between the employer and employee or the terms of a collective bargaining agreement.” 142 CONG. REC. 12234 (1996). Thus, the ECFA’s legislative history confirms its plain language: the “agreement” required by ECFA may be a condition of the employee’s employment.
2. The conditions Lojack placed on Rutti’s use of its vehicle did not make his commute compensable.
Rutti’s second argument is that restrictions placed on his use of the vehicle render the commute compensable. The ECFA provides that “activities performed by an employee which are incidental to the use of such vehicle for commuting shall not be considered part of the employee’s principal activities.” 29 U.S.C. [1053]*1053§ 254(a)(2). Rutti argues that the restrictions imposed by Lojack exceed the “incidental” and make his use of Lojack’s vehicle to commute an integral part of his principal activities for Lojack. Rutti cites Lojack’s restrictions against using the vehicle for personal pursuits and transporting passengers, the requirement that he drive directly from home to work and from work to home, and the requirement that he have his cell phone on. Rutti’s perspective finds no support in the language of the EOF A, is counter to its legislative history, and has been rejected by those courts that have considered the issue.
The legislative history shows that Congress recognized that employers would place conditions on their employees’ use of vehicles for commuting. House Report 585 commented that it “is not possible to define in all circumstances what specific tasks and activities would be considered ‘incidental’ to the use of an employers vehicle.” H.R.Rep. No. 104-585 at 5. However, it stated that communications between employer and employee, “routine vehicle safety inspections or other minor tasks, and transportation of tools and supplies, would not change the noncompensable nature of the travel.” Id. The minority report objected that as “non-employee passengers in such vehicles are uniformly prohibited,” an employee may be “effectively prohibited from engaging in the very common and often necessary family task of dropping off his or her child at school on the way to work.” Id. at 13. The failure of the minority report to stimulate any change in the bill indicates that Congress did not object to employers setting conditions on their employees use of company cars for commuting.
Those courts that have addressed this question have held that the cost of commuting is not compensable unless the employees show that they “perform additional legally cognizable work while driving to their workplace.” Adams, 471 F.3d at 1325; see also Smith v. Aztec Well Servicing Co., 462 F.3d 1274, 1286-87 (10th Cir. 2006) (noting that “[wjhile the Portal-to-Portal Act clearly excludes normal home to work travel from the scope of the FLSA, ... Congress ... still intended] for an employee’s activities to fall within the protection of the [FLSA] if they are an integral part of and are essential to the principal activities of the employees”) (quoting Steiner v. Mitchell, 350 U.S. 247, 254, 76 S.Ct. 330, 100 L.Ed. 267 (1956)) (internal quotation marks omitted).
The line between incidental and integral is well-illustrated by two cases from the Federal Circuit. In Bobo v. United States, 136 F.3d 1465 (Fed.Cir.1998), a group of Border Patrol agent dog handlers sought compensation for the time spent transporting their dogs between their homes and Border Patrol offices. Id. at 1466-67. They were not allowed to use the vehicles for personal use, were not allowed to make personal stops during their commute, were required to wear their official uniforms while using the vehicles, were required to monitor their radios, report their mileage and look out for suspicious activities. Id. at 1467. In addition, they were required “to make stops for the dogs to exercise and relieve themselves.” Id. Nonetheless, the Federal Circuit held that even accepting the restrictions as compulsory and for the benefit of their employer, “the burdens alleged are insufficient to pass the de minimis threshold.” Id. at 1468. The court specifically noted that “the main restriction on the INS Agents is the prohibition on rhaking personal stops during their commute)” and held that “such a restriction on their use of a government vehicle during their commuting time does not make this time compensable.” Id.
[1054]*1054Adams v. United States, 471 F.3d 1321 (Fed.Cir.2007), also concerned a suit by-government law enforcement agents seeking compensation for their commute from home to work in government-owned vehicles. Id. at 1323. They argued that they had to be available for emergency calls, had to have their weapons with them, had to monitor their communication equipment, could not run any personal errands, and had to proceed directly from home to work and back without unauthorized detours or stops. Id. The Federal Circuit held that pursuant to 29 U.S.C. § 254(a), merely commuting was insufficient; “the plaintiffs must perform additional legally cognizable work while driving to their workplace in order to compel compensation for the time spent driving.” Id. at 1325. The court further held that plaintiffs “had the burden of showing that their drive time was compensable work for FLSA purposes and of showing that it does not fall into the set of activities excluded from the definition of compensable work by the Portal-to-Portal Act as interpreted by our precedent.” Id. at 1326. The court concluded, citing its prior opinion in Bobo, that “[u]nder the Portal-to-Portal Act, plaintiffs’ driving time is not compensable.” Id. at 1327.
Here, Lojack placed fewer restrictions on Rutti’s use of its vehicle than were present in Adams and Bobo. More importantly, Rutti has failed to show that Lo-jack’s restrictions amount to “additional legally cognizable work.” Adams, 471 F.3d at 1325. The prohibition against carrying non-employee passengers was common practice before the statute was amended in 1996 and is not directly related to the “principal activities of the employees.” Aztec Well, 462 F.3d at 1287. In Bobo, the Federal Circuit specifically stated that the restriction on making personal stops did not make the commute time compensable. Bobo, 136 F.3d at 1468. Moreover, this restriction is not directly related to Rutti’s principal activities for Lojack. In addition, although the police officers in both Bobo and Adams were required to monitor their communications equipment, in neither case was this considered sufficient to compel compensation.7 In light of Rutti’s failure to cite any authority supporting a claim that Lojack’s restrictions constitute “additional legally cognizable work,” and because there are no material questions of fact as to the restrictions on Rutti’s use of Lojack’s vehicle, we affirm the district court’s determination that Rut-ti is not entitled under federal law to compensation for the time he spends commuting in Lojack’s vehicle.
B. One of Rutti’s off-the-clock activities may be compensable
Rutti also seeks compensation for activities that he engaged in for Lojack before he travels to his first job site and after he returns home from his last job site of the day. The ECFA, however, in addition to exempting commute time from compensation, also provides that an employer need not compensate an employee for “activities which are preliminary to or postliminary to said principal activity or activities.” 29 U.S.C. § 254(a)(2). Thus, to be entitled to compensation for his off-the-clock activities, Rutti must show that they are related to his “principal activities” for Lojack. In addition, our case law indicates that activity that might otherwise be compensable is not if the time involved is de minimis. Accordingly, we next discuss the applicable case law, first addressing the definitions of [1055]*1055“principal activities” and “de minimis ” for off-the-clock activities. We then apply those definitions to Rutti’s preliminary and postliminary activities.
1. Applicable Case Law
a. The definition of “principal activities ”
We first considered the definition of “principal activities” in Lindow v. United States, 738 F.2d 1057 (9th Cir.1984). The plaintiffs in that case sought overtime compensation for up to 15 minutes of work before the start of their shifts. Id. at 1059. We held that pre-shift activities are compensable if they are an “integral and indispensable part of the principal activities for which covered workmen are employed,” id. at 1060 (quoting Steiner v. Mitchell, 350 U.S. 247, 256, 76 S.Ct. 330, 100 L.Ed. 267 (1956)), and that the term “principal activities” is to be liberally construed “to include any work of consequence performed for an employer no matter when the work is performed.” Id. at 1061 (citing 29 C.F.R. § 790.8(a)).
The Fifth Circuit adopted a similar broad definition of principal activities in Dunlop v. City Elec., Inc., 527 F.2d 394 (5th Cir.1976). In Dunlop, electricians arrived at’ the work site approximately 15 minutes before the work day began at 8:00 a.m. to perform certain duties. Id. at 397. The issue was whether these activities were integral to their principal activities. The Fifth Circuit adopted a broad definition, holding that:
The test, therefore, to determine which activities are “principal” and which are “an integral and indispensable part” of such activities, is not whether the activities in question are uniquely related to the predominant activity of the business, but whether they are performed as part of the regular work of the employees in the ordinary course of business. It is thus irrelevant whether fueling and unloading trucks is “directly related” to the business of electrical wiring; what is important is that such work is necessary to the business and is performed by the employees, primarily for the benefit of the employer, in the ordinary course of that business. We find that the pre8:00 a.m. activities performed by the electricians and their helpers were within the broad range of “principal activities” performed at their employer’s behest and for the benefit of the business; as such they were compensable activities for which the employees would ordinarily have been paid had such work been performed during the normal workday.
Id. at 400^401 (footnote omitted).8
In Owens v. Local No. 169, Ass’n ofW. Pulp & Paper Workers, 971 F.2d 347 (9th Cir.1992), we approached “principal activities” from a different perspective. In Owens, the plaintiffs were mechanics who sought overtime compensation for the time that they were on call to receive emergency calls to fix equipment at their employer’s pulp mill. Id. at 348-49. We observed that the Supreme Court had held that time spent waiting for work is compensable if the waiting time is spent “primarily for the benefit of the employer and his business.” Id. at 350 (quoting Armour & Co. v. Wantock, 323 U.S. 126, 132, 65 S.Ct. 165, 89 L.Ed. 118 (1944)). We noted that “facts may show that the employee was ‘engaged to wait,’ which is compensable, or they may show that the employee ‘waited to be engaged,’ which is not compensable.” Id. at 350 (quoting Skidmore v. Swift & Co., 323 U.S. 134, 137, 65 S.Ct. 161, 89 L.Ed. 124 (1944)). We nonetheless [1056]*1056rejected the plaintiffs’ claims, holding that the proper test “is not the importance of on-call work to the employer, rather the test is focused on the employee and whether he is so restricted during on-call hours as to be effectively engaged to wait.” Id. at 354. See also Brigham v. Eugene Wafer & Elec. Bd., 357 F.3d 931, 938 (9th Cir.2004) (applying the Owens factors and vacating a grant of summary judgment for the employer because the factors narrowly favored the employees).
Thus, Lindow requires that we give “principal activities” a liberal construction “no matter when the work is performed,” 738 F.2d at 1061 (citing 29 C.F.R. § 790.8(a)). The Fifth Circuit’s opinion in Dunlop suggests that we pay particular attention to whether the activities “are performed as part of the regular work of the employees in the ordinary course of business.” 527 F.2d at 401. In addition, Owens counsels that we consider the extent to which the work impacts the employee’s freedom to engage in other activities. 971 F.2d at 354.
b. The definition ofde minimis
Our opinion in Lindow also applied the de minimis rule to claims of overtime compensation. 738 F.2d at 1062. Taking our lead from the Supreme Court’s opinion in Anderson v. Mt. Clemens Pottery Co., 328 U.S. 680, 692, 66 S.Ct. 1187, 90 L.Ed. 1515 (1946), we recognized that an employer’s obligation to pay for the employees’ efforts had to be moderated by a de minimis rule.9 We cited the Supreme Court’s statement that:
When the matter in issue concerns only a few seconds or minutes of work beyond the scheduled working hours, such trifles may be disregarded. Split-second absurdities are not justified by the actualities of working conditions or by the policy of the Fair Labor Standards Act. It is only when an employee is required to give up a substantial measure of his time and effort that compensable working time is involved.
Lindow, 738 F.2d at 1062 (quoting Anderson, 328 U.S. at 692, 66 S.Ct. 1187).
In Lindow, we concluded that even though some of the employees’ activities were principal activities, the district court properly determined that “the 7 to 8 minutes spent by employees reading the log book and exchanging information, even if not preliminary, was de minimis and therefore not compensable.” Id. at 1062.
The panel carefully explained its reasoning. An important factor in determining whether a claim is de minimis is the amount of time spent on the additional work. The panel specifically stated that: “[tjhere is no precise amount of time that may be denied compensation as de minim-is. No rigid rule can be applied with mathematical certainty.” Id. at 1062. Nonetheless, the panel noted that most courts “have found daily periods of approximately 10 minutes de minimis even though otherwise compensable.” Id. The panel then qualified this statement, ex[1057]*1057plaining: “[t]he de minimis rule is concerned with the practical administrative difficulty of recording small amounts of time for payroll purposes.” Id. Accordingly, employers “must compensate employees for even small amounts of daily time unless that time is so minuscule that it cannot, as an administrative matter, be recorded for payroll purposes.” Id. at 1062-63. Furthermore, courts may consider “the size of the aggregate claim,” and “have granted relief for claims that might have been minimal on a daily basis but, when aggregated, amounted to a substantial claim.” Id. at 1063. Also, courts apply “the de minimis rule in relation to the total sum or claim involved in the litigation.” Id. “Finally, in applying the de minimis rule, we will consider whether the claimants performed the work on a regular basis.” Id. The court summarized its position as follows:
in determining whether otherwise compensable time is de minimis, we will consider (1) the practical administrative difficulty of recording the additional time; (2) the aggregate amount of compensable time; and (3) the regularity of the additional work.
Id. at 1063.
Thus, in determining whether an otherwise compensable activity is de minimis, we apply the three-prong test set forth in Lindow. In doing so, we recognize that the test reflects a balance between requiring an employer to pay for activities it requires of its employees and the need to avoid “split-second absurdities” that “are not justified by the actuality of the working conditions.” Lindow, 738 F.2d at 1062 (quoting Anderson, 328 U.S. at 692, 66 S.Ct. 1187).10
2. Applying the applicable law to Rut-ti’s off-the clock activities
Rutti’s off-the-clock activities may be divided between those that take place before he leaves home, i.e. his preliminary activities, and those that take place after he returns home, i.e. his postliminary activities.
a. Preliminary Activities
Rutti’s morning activities do not appear to be integral to his principal activities. Most of his activities — “receiving, mapping, and prioritizing jobs and routes for assignment”- — -are related to his commute. Under the FLSA, commuting is presumptively noncompensable, and is clearly distinct from Rutti’s principal activities for Lojack. Although there are some indications that Rutti also filled out some forms for his jobs at home, it is not clear that the paperwork could not be performed after Rutti reached the job site, or that Lojack required the forms to be filled out before Rutti reached the job site.
In any event, these preliminary activities, to the extent that they are both distinct from his commute (which is not compensable) and related to his principal activities, appear to be de minimis, and [1058]*1058thus, not compensable. Even though Rutti allegedly filled out certain forms every morning, there is nothing to suggest that this took more than a minute or so. Thus, viewing the facts in the light most favorable to Rutti, he has not offered any evidence of preliminary activities that are both integral to his principal activities for Lojack and take more than a de minimis amount of time. Accordingly, the district court properly granted Lojack summary judgment on Rutti’s claim for compensation for preliminary activities.
b. Postliminary Activities
Lojack requires that Rutti, after he completes his last job for the day and goes “off-the-clock,” return home and send a PDT transmission to Lojack using a modem provided by Lojack. The transmissions have to be made every day as they provide Lojack with information concerning all the jobs its technicians perform during the day. The transmissions appear to be “part of the regular work of the employees in the ordinary course of business,” and are “necessary to the business and [are] performed by the employees, primarily for the benefit of the employer, in the ordinary course of that business.” Dunlop, 527 F.2d at 401. Accordingly, at least on summary judgment, the district court could not determine that this activity was not integral to the Rutti’s principal activities.
Lojack might still be entitled to summary judgment, if it could be determined that this postliminary activity was clearly de minimis. The evidence before the district court, however, does not compel such a conclusion. The fact that several technicians testified that they spent no more than five to ten minutes a night on PDT transmissions might appear to give rise to a presumption that an activity is de minimis, see Lindow, 738 F.2d at 1062, but such a conclusion is neither factually nor legally compelling.
It is not factually compelling because, although it may take only five to ten minutes to initiate and send the PDT transmission, the record shows that the employee is required to come back and check to see that the transmission was successful, and if not, send it again. There is also evidence in the record that there are frequent transmission failures. Accordingly, the record does not compel a finding that the daily transmission of the record of the day’s jobs takes less than ten minutes.
Furthermore, we have not adopted a ten or fifteen minute de minimis rule. Although we noted in Lindow, that “most courts have found daily periods of approximately 10 minutes de minimis even though otherwise compensable,” we went on to hold that “[t]here is no precise amount of time that may be denied compensation as de minimis ” and that “[n]o rigid rule can be applied with mathematical certainty.” 738 F.2d at 1062. The panel went on to set forth a three-prong standard, which would have been unnecessary if the panel had intended to adopt a ten or fifteen minute rule.11
The application of this three-prong test to the facts in this case do not compel a conclusion that the PDT transmissions are de minimis. The first prong, “the practical administrative difficulty of recording the additional time,” id. at 1063, is closely balanced in this case. Certainly, it is difficult to determine exactly how much time each technician spends daily on the PDT transmissions. It is also not clear what [1059]*1059activities should be covered. Is the time when the technician comes back to check to see if the transmission was successful included? When a technician is waiting until ten minutes after the hour, is he “engaged to wait” or “waiting to be engaged?” See Owens, 971 F.2d at 350. Although it may be difficult to determine the actual time a technician takes to complete the PDT transmissions, it may be possible to reasonably determine or estimate the average time. For example, there is evidence in the record that Lojack had agreed to pay one technician an extra 15 minutes a day to cover the time spent on PDT transmissions. In sum, the inherent difficulty of recording the actual time spent on a particular PDT transmission does not necessarily bar a determination that the PDT transmissions are not de minimis. See Reich v. Monfort, Inc., 144 F.3d 1329, 1334 (10th Cir.1998) (holding that the time it took meat packers to don and shed their employer-mandated clothing was not de minimis even though “the practical difficulty of supervising and recording the additional time weighs in favor of finding it noneompensable”).
The other two prongs, “the aggregate amount of compensable time,” and “the regularity of the additional work,” Lindow, 738 F.2d at 1063, favor Rutti. Rutti asserts that the transmissions take about 15 minutes a day. This is over an hour a week. For many employees, this is a significant amount of time and money. Also, the transmissions must be made at the end of every work day, and appear to be a requirement of a technician’s employment. This suggests that the transmission “are performed as part of the regular work of the employees in the ordinary course of business,” Dunlop, 527 F.2d at 401, and accordingly, unless the amount of time approaches what the Supreme Court termed “split-second absurdities,” the technician should be compensated. See Anderson, 328 U.S. at 692, 66 S.Ct. 1187.
Our review of the record suggests that the PDT transmissions are an integral part of Rutti’s principal activities and that there are material issues of fact as to whether the PDT transmissions are de minimis. Accordingly, the grant of summary judgment in favor of Lojack on Rut-ti’s claim for the transmissions must be vacated. See Balint v. Carson City, Nev., 180 F.3d 1047, 1054 (9th Cir.1999) (holding that in reviewing a grant of summary judgment, we do “not weigh the evidence or determine the truth of the matter, but only determines whether there is a genuine issue for trial”). This does not mean that on remand, Lojack may not be able to make a persuasive factual showing for summary judgment under the standard clarified in this opinion. We, however, decline to make such a decision in the first instance.
C. Rutti’s off-the-clock activities do not extend his workday under the continuous workday doctrine.
Finally, Rutti argues that under the continuous workday doctrine,12 because his work begins and ends at home, he is entitled to compensation for his travel time, citing Dooley v. Liberty Mutual Ins. Co., 307 F.Supp.2d 234 (D.Mass.2004). In Dooley, automobile damage appraisers sought compensation for the time they spent traveling from their offices in their homes to locations where they inspected [1060]*1060damaged cars. Id. at 239. The district court first determined that the work the appraisers undertook at home constituted principal activities.13 Id. at 242. The court then determined that compensation was not prohibited by the Portal-to-Portal Act,14 and concluded that those appraisers who could show that they performed work at home before or after their daily appraisals were entitled to compensation. Id. at 249.
Even were we to adopt the continuous workday doctrine set forth in Dooley, Rut-ti would not be entitled to compensation for his travel time to and from the job sites. We have already determined that Rutti’s preliminary activities that are not related to his commute are either not principal activities or are de minimis. Accordingly, his situation is not analogous to the situation in Dooley. See 307 F.Supp.2d at 245 (“The first and last trip of the day for these appraisers is not a commute in the ordinary sense of the word-it is a trip between their office, where their administrative work is performed, and an off-site location.”).
Our determination that Rutti’s postliminary activity, the PDT transmission, is integrally related to Rutti’s principal activities might support the extension of his work day through his travel back to his residence, were it not for 29 C.F.R. § 785.16. This regulation provides that “[pjeriods during which an employee is completely relieved from duty and which are long enough to enable him to use the time effectively for his own purposes are not hours worked.”15 Lojack allows a technician to make the transmissions at any time between 7:00 p.m. and 7:00 a.m. Thus, from the moment a technician completes his last installation of the day, he “is completely relieved from duty.” His only restriction is that sometime during the night he must complete the PDT transmission. Because he has hours, not minutes, in which to complete this task, the intervening time is “long enough to enable him to use the time effectively for his own purpose.” See Mireles v. Frio Foods, Inc., 899 F.2d 1407, 1413 (5th Cir.1990) (holding that waiting time “greater than forty-five minutes are not compensable because [1061]*1061Plaintiffs were not required to remain on Defendant’s premises during such periods and could use such periods effectively for their own purposes”). Rutti has not shown that the district court erred in determining that neither his preliminary nor postliminary activities extended his workday under the continuous workday doctrine.
IV.
We substantially agree with the district court’s grant of summary judgment in favor of Lojack on Rutti’s claims for overtime compensation. We agree that Rutti is not entitled to compensation under federal law for his commute to and from his job sites in a vehicle provided by Lojack. We agree that Rutti is not entitled to compensation for his preliminary activities as those activities are either not principal activities, or if principal activities, are de minimis. However, we vacate the district court’s grant of summary judgment on Rutti’s claim for compensation for the PDT transmissions because it appears that this function is integral to Rutti’s principal activities for Lojack and the record does not compel a determination that the time consumed by this function is de minimis.
We also agree with the district court that Rutti is not entitled to compensation for his commute time under the continuous workday doctrine because, as noted, his preliminary activities are either not principal activities or are de minimis. In addition, accepting that the postliminary PDT transmission is a principal activity, the fact that Rutti is not required to make the transmission at any specific time means that he may use the intervening time for his own purposes, and accordingly pursuant to 29 C.F.R. § 785.16, Lojack need not compensate him for the intervening time.
The district court’s grant of summary judgment is affirmed except as to Rutti’s claims for compensation for commuting under state law and for the required postliminary PDT transmission, which are vacated. This matter is remanded for further proceedings consistent with this opinion. Each party to bear its own costs.
AFFIRMED in part, VACATED in part, and REMANDED.
Separate Opinion by Judge SILVERMAN.1
Compensation for Commute Time under California Law
Rutti contends that even if his commute is not compensable under the federal Employment Commuter Flexibility Act, it is nevertheless compensable under California law. California law requires that employees be compensated for all time “during which an employee is subject to the control of an employer.” Morillion v. Royal Packing Co., 22 Cal.4th 575, 578, 94 Cal.Rptr.2d 3, 995 P.2d 139 (2000). In Morillion, the California Supreme Court held that the plaintiffs were “subject to the control” of their employer during a mandatory bus commute because “plaintiffs could not drop off their children at school, stop for breakfast before work, or run other errands requiring the use of a ear.” Id. at 586, 94 Cal.Rptr.2d 3, 995 P.2d 139. The California Supreme Court reasoned that the “[plaintiffs were foreclosed from numerous activities in which they might otherwise engage if they were permitted to travel to the fields by their own transportation.” Id. That is precisely the situation here. Rutti was required to drive the company vehicle, could not stop off for personal errands, could not take passengers, was required to drive the vehicle directly from home to his job and back, and could not use his cell phone while driving except that he had to keep his [1062]*1062phone on to answer calls from the company dispatcher. In addition, Lojack’s computerized scheduling system dictated Rut-ti’s first assignment of the day and the order in which he was to complete the day’s jobs. There is simply no denying that Rutti was under Lojack’s control while driving the Lojack vehicle en route to the first Lojack job of the day and on his way home at the end of the day.
The dissent attempts to distinguish Morillion by summarily concluding that “Rutti’s use of Lojack’s automobile to commute to and from his job sites is more analogous to the ‘home to departure points’ transportation in Morillion than to the employees’ transportation on the employer’s buses.” Aside from the lack of factual analysis to support this ipse dixit, the dissent also utterly ignores the relevant question under California law, which is whether Rutti was “subject to the control of an employer” during his mandatory travel time. A straightforward application of Morillion easily answers that question in the affirmative. Rutti was required not only to drive the Lojack vehicle to and from the job site, but was forbidden from attending to any personal business along the way. Because he was obviously under the employer’s control in these circumstances he was, under California law, entitled to be paid.2
The dissent makes the mistake of assuming that any employer-mandated travel that begins or ends at home is automatically noncompensable, but that assumption again ignores the controlling legal principle. Under California law it is the “level of the employer’s control over its employees” that “is determinative,” not whether the employee just so happens to depart from, or return to, his home instead of some other location. Id. at 587, 94 Cal. Rptr.2d 3, 995 P.2d 139. Here, the level is total control. To repeat, Rutti was required to use the company truck and was permitted no personal stops or any other personal use. Thus, under Morillion, Rutti had a valid state-law claim for compensation. The district court’s grant of summary judgment as to Rutti’s claim for compensation for commuting under California law is reversed.