Evaluation of Auxiliary Signing for Driveway Assistance Devices at Low-Volume Access Points within One-Lane,Bidirectional Work Zones

Abstract

Work zone traffic control presents unique challenges during single-lane closures on two-lane, two-way facilities. In these circumstances, traffic regulators (i.e., flaggers or temporary signals) ensure alternating one-way travel in the work zone. However, the use of traffic regulators becomes less practical at mid-segment low-volume access points, such as driveways and minor intersections, within the lane closure. These locations present additional risk for flaggers and, owing to financial infeasibility, are often rendered without traffic control. A recently developed experimental traffic control treatment, the driveway assistance device (DAD), is designed to provide guidance to motorists for a safe entrance into the lane closure from low-volume access points. As this device is relatively new, much is still unknown about the optimal device design and operation to provide the highest compliance. To address this, a field study was performed to evaluate driver compliance associated with DADs implemented at a series of driveways along a one-lane, two-way work zone along a state highway in northern Michigan. As little guidance exists for DAD auxiliary signage, the field study specifically sought to compare the effects of five different sign messages utilized with the DAD. The driver compliance data revealed that the most effective sign messages were those that included “Turn” instead of “Yield,” a prominent “WAIT” message at the top, in addition to a supplemental “No Turn on Red” sign. Based on these findings, DADs are recommended for continued experimental use along with appropriate auxiliary signage at work zones where one-lane, two-way traffic is being maintained.

Keywords

operations MUTCD safety traffic signals traffic signs work zones

The National Safety Council reported that between 2011 and 2021, a total of 607 road workers were struck and killed by motor vehicles, ranging from 44 to 65 fatalities annually ( 1 ). To help reduce worker injuries and fatalities, recently developed traffic control devices aim to regulate traffic flow while simultaneously lowering the risk of fatalities and injuries to road workers. One particularly challenging work zone traffic control situation arises along single-lane closures on two-lane, two-way roads. In these situations, traffic regulators are employed to alternate one-way traffic flow on the facility. Often this is accomplished by stationing flaggers at both ends of the work zone and at major access points. To reduce worker exposure and subsequent crash risk, portable traffic control signals (PTCSs) or automatic flagger assistance devices (AFADs) are often utilized in these situations ( 2 , 3 ). However, the use of traffic regulators becomes less practical at mid-segment low-volume access points within work zone lane closures. These locations present additional risk for traffic regulators and, owing to financial infeasibility, are often rendered without traffic control ( 2 , 3 ). This creates a situation in which drivers exiting low-volume access points are not provided with the necessary information to judge the current direction of the work zone traffic flow.

The driveway assistance device (DAD) has recently been developed to provide traffic control to low-volume access points on a one-lane, two-way work zone without the need for a flagger. The device was initially developed and refined by the Texas Department of Transportation (TxDOT). DADs are equipped with alternating left and right flashing arrows to indicate current traffic flow direction, and a steady red signal to indicate the stop condition ( 4 ). DADs are wirelessly interconnected with PTCSs at both ends of the work zone in a coordinated timing plan. Various device configurations (e.g., doghouse, horizontal, red-over-yellow) and arrow colors (red, yellow) have been employed in experiments across the country. Often, the device is deployed with auxiliary signage and multiple state agencies install a “No Turn on Red” (NTOR) sign. An example of a DAD employed on a one-lane, two-way work zone is presented in Figure 1.

Figure 1.

Example driveway assistance device (DAD).

Overall, the innovative nature of the DAD means that, to date, there is little body of work on best practices related to the device. Initial research indicates a high potential for legal and/or safe movement rates, but also issues related to motorist comprehension ( 4 , 5 ). This, along with the variability in DAD design across states, has hampered device adoption into the Manual on Uniform Traffic Control Devices for Streets and Highways (MUTCD). Thus, the DAD remains experimental and requires formal Federal Highway Administration (FHWA) approval for usage. To this end, additional research is necessary to determine best usage and design practices for the DAD and develop guidelines for its use ( 4 – 6 ).

The objective of the current research was to evaluate DAD display and auxiliary sign messages that most effectively and safely direct motorists to enter the mainline of a one-lane, two-way work zone, and ultimately to assist with the eventual adoption of the device by the FHWA and inclusion within the MUTCD. Several auxiliary signing strategies were developed and tested in a field study in northern Michigan where DADs were deployed at low-volume access points along a one-lane, two-way work zone. In this field study, the DADs were tested alongside various auxiliary signs to evaluate motorist compliance related to the device with respect to the sign message.

Literature Review

Existing Guidelines and Research for One-Lane, Two-Way Work Zones

Chapter 6C of MUTCD outlines temporary traffic control elements ( 7 ). Section 6C.10 states, “provisions should be made for alternate one-way movement … via methods such as flagger control, a flag transfer, pilot car, traffic control signals, or stop or yield control.” The manual also states that when neither ends of a work zone are visible, flagging procedures or PTCSs, with or without a pilot car, should be used.

Both MUTCD and the Michigan MUTCD offer guidance on using PTCSs and AFADs in a work zone ( 7 , 8 ). However, there is no specific language related to a DAD. In both manuals, chapters 6E.04–.07 have guidelines for using AFADs at the end of a one-lane, two-way work zone, stating that when using an AFAD, there must either be a traffic regulator in the work zone that controls both devices, or one controlling each. The manuals require that an AFAD has incorporated gates or bars to block the stopped direction of traffic. Section 6E.06 states than an AFAD should have a “Stop Here on Red” (R10-6[a]) sign. Furthermore, the guidelines only mention a flashing yellow and the inclusion of a change interval between this and a solid red phase.

A study by Schrock et al. evaluated the effectiveness of three mainline work zone traffic control treatments, PTCSs only, flagger only, and PTCSs with flagger ( 9 ). They found that PTCSs had a 3.1% violation rate—higher than the 1.1% for flagger only and 1.3% for the combination treatment. This study also incorporated an investigation into operational efficiencies of the treatments and found that delays decreased by 5% with the presence of a flagger. A related study by Schrock et al. evaluated the appropriate traffic characteristics for PTCSs with a pilot car ( 10 ). These results found that a PTCS system would fail above an annual average daily traffic of 7,083, given a maximum delay threshold of 15 min and maximum pilot car speed of 40 mph. The optimum maximal green time in this case was 446 s. Later research by Finley and Theiss assessed compliance with mainline PTCSs with pilot vehicle treatments, both with and without flaggers ( 11 ). The researchers argued that the necessity of flagger presence negated the safety and construction operational benefits that should come with use of a PTCS and pilot vehicle. The results showed that 2.7% and 2.3% of motorists were noncompliant, with and without flaggers, respectively, and the difference between the two rates was not statistically significant. Given the high compliance rates, the authors of the study suggested a MUTCD revision to allow for the use of PTCSs at one-lane, two-way work zones without a flagger ( 11 ).

The previously mentioned study by TxDOT that developed DADs used motorist surveys to investigate comprehension of two AFAD designs ( 4 ). For both device designs, participants were shown a sequence of videos depicting progression of the device in use. All participants understood that, for a red-yellow device, a steady red signal indicated a stop. Furthermore, in all but one of seven scenarios, the results showed that at least 94% understood a flashing yellow arrow signal to mean proceed with caution. This result was consistent whether participants were shown a stop, proceed, or transition phase, or if the transition phase was omitted. For the slow-stop device, participant actions were recorded in response to varying auxiliary signs. The most effective sign used phrasing with symbolic depictions of a “Stop and Slow” sign to provide guidance. With this sign, 92% of participants indicated they would remain stopped until a slow sign was shown. However, this was not significantly different from the 85% comprehension threshold. A synthesis of the results from both device designs could lend weight to using a red/flashing-red configuration for this study, since “Proceed with Caution” with flashing yellow arrows is not the preferred indication for Michigan usage. The results of the application of auxiliary signs indicated the need for further investigation of simple, effective, and symbolic signs alongside a DAD.

Driveway Assistance Devices

The same study by TxDOT used a modified hybrid (doghouse) device and a disappearing legend symbolic device to conduct a field study at a rural intersection work zone in Texas ( 4 ). Motorists, both with and without the presence of a researcher in the vehicle, were observed to evaluate device comprehension. The results for the motorists absent a researcher found that 87% and 92%, respectively, understood the doghouse and symbolic devices. The misinterpretations of the device were attributed to the flashing yellow arrows that were more easily confused than the symbolic nature of the disappearing legend sign. The authors acknowledged that the disappearing legend device is more expensive and less readily available. Similarly, it was noted that further research would be necessary to improve motorist comprehension of both devices.

The research evaluation plan for a later TxDOT project discussed DAD usage around the country ( 6 ). Overall, the plan indicated a lack of standardization in DAD design between states. Within this report, the New Jersey DAD (horizontal configuration) is described as the most widely employed. The study makes recommendations on future configuration options for DADs. A strong recommendation is made to experiment with the doghouse design and to include an NTOR sign to improve comprehension. Concerns were raised that many states, including Michigan, use flashing red arrows, which indicate the requirement to stop before proceeding. Thus, TxDOT recommended to continue DAD experimentation with flashing yellow arrows. Additional guidance states that it would be beneficial for further research on auxiliary signs used with DADs.

A Michigan Department of Transportation (MDOT) FHWA experimentation final report summarized the results of five pilot tests of horizontal DADs at residential driveways within the state between 2015 and 2018 ( 5 ). The combined rate of correct movements between all tests was 82.8%., It was also found that another 15.7% of motorists proceeded incorrectly, but safely, indicating an overall safe movement rate of 98.5%. MDOT used these legal and safe movement rates to conclude that the device is safe and effective at regulating traffic at residential driveways in a one-lane, two-way work zone, and recommended that the device may be considered in the upcoming iteration of MUTCD. Although the pilot tests were a success overall, MDOT’s final report noted multiple concerns with device operation. The report notes the importance of work zone visibility, geometrics, and volume. A supplementary recommendation indicates that provision of left- and right-turn lanes at certain access points may relieve excessive queues. This suggestion was made in conjunction with the need for further research at higher-volume access points.

Moreover, there are multiple projects nationwide that are currently investigating DADs. An ongoing study in Nebraska ( 12 ) has similar goals to the MDOT pilot studies, while also incorporating device usage at stop-controlled locations. Furthermore, TxDOT continues to evaluate DADs in a current project. DADs have also been approved for experimentation by state transportation agencies in Indiana, Massachusetts, Mississippi, New York, and Vermont ( 13 ).

Field Study

A field study was conducted to evaluate various sign messages that were selected based on a combination of state agency practice along with input from the project technical panel. The field study was carried out at a work zone along US-31 in Benzonia, MI during June, July, and August of 2020 and is explained in greater detail in the final project report ( 14 ). Under normal circumstances, this segment of US-31 has a two-lane cross-section with a two-way left-turn lane. The project limits extended from the US-31 at M-115 intersection to the north to approximately 800-ft south of the bridge over the Betsie River, for a length of nearly 1.5 mi. The work included full-depth hot mix asphalt reconstruction, curb and gutter removal and replacement, and drainage improvements. Figure 2 depicts the US-31 work zone utilized for this field study.

Figure 2.

US-31 work zone near Benzonia, MI used for field evaluation.

The research team collaborated with MDOT to create a plan in which the DADs could be implemented along the work zone. The construction was performed in three stages, working from south to north and passing numerous low-volume access points. The traffic control strategy for US-31 was to be a single-lane closure with alternating one-way traffic. For each stage, PTCSs were placed at the ends of the work zones to regulate the mainline. The staging of this project allowed for four DADs to be placed at numerous low-volume access points during the project.

DAD Configuration and Placement

The DADs were placed directly in the line of sight of drivers exiting low-volume access points. Variability in the access point geometry resulted in some DADs being placed at the end of the driveway, or directly across from the driveway on the far side of the intersection. All DADs used in this study were of the horizontal device configuration, which is the most common DAD signal configuration ( 6 ) and the standard utilized in Michigan. Each device had a 12-in. red signal in the center, along with left and right red arrows indications that were either 8- or 12-in diameter. In collaboration with MDOT, red arrows were selected for this project. Red arrows had been used in the prior MDOT study in which it was noted that the connotations of red arrows gave a stronger safety message (i.e., stop) than yellow. Each DAD assembly was generally raised to 7 ft to ensure adequate motorist visibility, with minor adjustments related to access point lines of sight. The devices were connected and operated in synchrony with the PTCSs at the ends of the work zone on the mainline. The flashing red arrow phase was sequenced with the PTCSs green to allow drivers to access the mainline before or partially during platoon arrival. The flashing red arrow phase was terminated at the end of the mainline green, at which point the steady circular red signal became illuminated. The steady circular red indication remained, as a clearance phase, until the green phase for the opposite traffic flow direction. At this point the associated opposite arrow phase would begin. The timing was adjusted based on the distance between PTCSs. The setup in Figure 3 is typical of the DADs utilized in the field study. Alongside each DAD assembly was a portable mount where auxiliary signs were placed. The NTOR sign was present at each access point and every auxiliary sign was visible to motorists exiting the access points. Oncoming and waiting opposing platoons were visible to access point drivers throughout the study, owing to the configuration and length of the work zone staging.

Figure 3.

Typical DAD and auxiliary signage used in the field evaluation.

Video Recording

Data on driver compliance were collected using video cameras that simultaneously viewed the access point, mainline, and DAD indication. The video was captured using commercially available high-definition video cameras. The cameras were mounted to telescoping aluminum poles that could be easily raised depending on site characteristics. Pole assemblies and small video cameras facilitated discreet recording, so that the presence of the camera was concealed at both the access point and the mainline. Camera setup took less than 5 min and was facilitated without assistance from MDOT. Each recording session occurred in clear, dry, daylight conditions that allowed for two 3-h camera sessions per site on a recording day. An example screenshot from the video data is presented in Figure 4.

Figure 4.

Example screenshot from video at DAD access point.

Experimental Auxiliary Signs

A series of five auxiliary signs were deployed during the field study, which are shown in Figure 5. All auxiliary signs were mounted alongside the DAD and an NTOR sign. The sign messages were selected by the technical panel, and the signs were produced by the traffic control contractor for the work zone. The “Proceed” sign was selected as it is the standard DAD signage used by MDOT. FHWA had expressed interest in testing the “Turn Only in Direction of Arrow” message. The “Stop and Wait for Arrow before Turning” sign is a variation of a sign used alongside DADs in Nebraska. MDOT’s expression of interest in the “WAIT” message is incorporated in the final two signs.

Figure 5.

Auxiliary signs used in the field study: (a) “Proceed on Flashing Red Arrow after Stop”, (b) “Turn Only in Direction of Arrow”, (c) “Stop and Wait for Arrow before Turning”, (d) “WAIT for Arrow before Turning”, and (e) “WAIT Turn Only in Direction of Arrow”.

Data Collection

Data were collected at 10 low-volume access points along the US-31 work zone. Each location netted at least one full day of data collection, however, construction phasing and low traffic volumes resulted in the infeasibility of testing all five signs at each access point. At the two highest-volume sites, McDonald’s and Save-a-Lot, a rotation of all five auxiliary signs was tested over the course of a 5-day period in August 2020. Data were collected at both sites across all 5 days in a morning and afternoon session. This rotation allowed for two signs to be tested at one site on a given day. The layout of the sign rotation for the field study is further detailed in Figure 6.

Figure 6.

DADs study sites and test signs.

Video data were extracted by trained graduate research assistants in a manual video review process. A data collection spreadsheet was designed so that relevant data could be collected for each vehicle observed to exit a DAD-controlled access point. Frame-by-frame time stamp data were collected using Quicktime V7.7.9 software. When a vehicle arrived in the camera view, a time stamp and its queue position were recorded. The DAD indication and time stamp were noted on the vehicle reaching the front of the queue and on the vehicle entering the mainline. Additional information was collected for each vehicle, including turn direction and legality (given the device indication) at the time of movement. Example scenarios of safe and legal vehicular movements are provided in Table 1.

Table 1.

Example Scenarios for Safe and Legal Vehicular Movements

Safe/unsafe	Scenario
Legal
Safe	Right turn on a right arrow
Safe	Left turn on a left arrow
Unsafe	Driving around queue leader to enter mainline
	Accepting a short gap and forcing mainline to slow for exiting vehicle
	Driving into the closed lane before entering mainline
Illegal
Safe	Joining end of platoon on a steady solid red indication
Safe	Accepting a gap within the platoon during steady red indication
Unsafe	Entrance in the direction opposite of arrow indication
Unsafe	Driving into oncoming traffic

Based on the conditions listed above, the research assistants recorded each observed movement as safe or unsafe. When the vehicle reached the front of the queue, the observer noted whether it came to a full stop before entering the mainline. The final data point recorded whether a vehicle joined a mainline platoon or entered the mainline before the arrival of the platoon. Queue, dwell, and gap times were calculated via the use of time stamps recorded throughout the vehicles’ observed DAD interaction. Queue time was defined as the time a vehicle spent on camera before becoming the leader of the queue. Dwell time was the time each vehicle spent as queue leader before entering the mainline. Gap time was the time between a vehicle’s entrance to the mainline and the following vehicle in the mainline platoon.

Data Summary and Preliminary Analysis

Overall, data collected from the field study comprised 2,258 vehicle observations considering all study access points and auxiliary sign combinations. Table 2 presents a summary of legal movement rates across all sign and site combinations. The data showed that, considering all sites and signs, the legal movement rate was 62%. This rate varied across sites mostly between 47% and 66%, except for the lowest volume site, a minor intersection. As a comparison of sign types, the legal movement rate ranged between 57% and 69%. Legal movement rates in this field study were lower than prior studies ( 4 , 5 ). This phenomenon could be associated with longer wait times as a function of higher volumes at these test sites compared with prior studies.

Table 2.

Total Observations and Legal Movement Rates for Vehicles Exiting a DAD Driveway, by Location and Sign

Site	Sign message						Total (legal %)
Site	Proceed	Turn only	Stop and wait	WAIT for arrow	WAIT turn only	No sign	Total (legal %)
McDonald’s	298	103	351	285	209	x	1,246 (66)
Save-a-Lot	92	63	63	53	59	x	330 (55)
Trailer park	76	86	x	x	88	x	250 (64)
State Savings Bank	139	x	x	x	x	x	139 (53)
Dispensary	30	x	x	x	32	x	62 (58)
South St. EB	54	x	x	x	x	x	54 (63)
South St. WB	36	x	x	x	x	x	36 (50)
Walker St. EB	55	x	x	x	x	x	55 (47)
Walker St. WB	17	x	x	x	x	x	17 (12)
River St. EB	x	x	x	x	x	69	69 (57)
Total (legal %)	797 (57)	252 (60)	414 (63)	338 (67)	388 (69)	69 (57)	2,258 (62)

Note: EB = eastbound; WB = west bound; DAD = driveway assistance device.

“x” indicates the sign was not used.

The field study results also indicated promising levels of safety with 2,094 safe movements overall, comprising 93% of all observations. However, there were inherent incongruities with classifying the safety of a movement, as there were inherent differences in test sites related to construction work activities and temporary traffic control devices. Given the circumstantial situations of safe or unsafe observations, and whether they could be associated with DAD or auxiliary sign design, no formal analysis of movement safety was conducted. However, the 93% safe movement rate is a promising aspect of device performance in the field.

One of the primary objectives of this study was to determine the influence of sign configuration on legal movements for vehicles interacting with a DAD. The full five-sign rotation was conducted at only two sites, McDonald’s and Save-a-Lot. However, only the recordings at McDonald’s yielded an adequate data sample, comprising 1,246 observations; the decision was therefore made to limit the analysis of legal movement to these observations. The McDonald’s averaged 66% legal and 93% safe movements across all sign combinations. Crossing movements at the McDonald’s were excluded from analysis owing to the current DAD lacking provision for a legal through indication. This resulted in 1,221 observations being usable for the analysis of legal movement rates. The frequencies and percentages of categorical outcomes of both independent and dependent variables, including a breakdown of legal movement rates by sign for all McDonald’s observations, are presented in Table 3.

Table 3.

Descriptive Statistics for Vehicles Exiting McDonald’s Driveway, by DAD Sign Treatment

Total observations and legal movement rate for vehicles exiting McDonald’s driveway, by sign
Actions						Total
Total exiting vehicle observations	275	103	350	285	208	1,221
Percentage legal movements	63	64	65	72	73	69
Percentage safe movements	92	92	92	93	98	93
Frequency and percentage of independent variable categories
Variable categories		Frequency	Percentage	Variable categories	Frequency	Percentage
Queue position = 1		634	52	DAD display = solid	467	38
Queue position = 2		273	22	DAD display = right	493	40
Queue position = 3		314	26	DAD display = left	261	22
Full stop = yes		640	52	Vehicle type = car	261	22
Full stop = no		581	48	Vehicle type = SUV	535	44
Vehicle direction = left		415	34	Vehicle type = van	115	9
Vehicle direction = right		806	66	Vehicle type = pickup	295	24
Platoon = precede		818	67	Vehicle type = other	15	1
Platoon = join		403	33

Note: DAD = driveway assistance device; SUV = sports utility vehicle.

Binary Logistic Regression Analysis

The dependent variable for this analysis was the legality of movements for each vehicle observed to exit the McDonald’s driveway. Movement legality was tested as a function of the test sign and other factors related to the vehicle, site, and scenario. Discrete outcome models are an appropriate framework for this analysis because of the binary dependent variable (i.e., legal versus illegal movements). The binary logistic regression model was applied, given its use in predicting the probability of a dichotomous outcome from a set of predictor variables. Binary logistic regression was selected for this analysis given its similarity to linear regression and its advantage of a categorical response variable. Models from the binary logistic family take the following form:

Y_{i} = logit (P_{i}) = \ln (\frac{P_{i}}{1 - P_{i}}) = β_{0} + β_{1} X_{i 1} + β_{2} X_{i 2} + \dots + β_{k} X_{ik}

(1)

where the response variable, Y_i, is logistic transformation of the probability of a legal movement for vehicle i, which is also denoted as P_i. Like a linear regression framework, the independent variables are indicated as X_i1 to X_ik and affect the probability of a vehicle’s legal movement. The regression intercept is given as $β_{0}$ , whereas each regression coefficient is given as $β_{1}$ through $β_{k}$ . The estimated odds ratio from this regression is then interpreted as the percent change in likelihood as a result of a unit change in any of the variables included in the model.

Categorical independent variables were coded as a set of n-1 discrete binary variables to be assessed relative to a designated base condition. The five sign conditions included in the model are those shown in Figure 5. The “Proceed” sign was selected as the baseline condition for this, as it is the standard sign used by MDOT alongside all DAD installations. This allowed for the sign-related parameter estimates to be interpreted as the difference between the standard MDOT sign and the alternative test sign. Gap-, dwell-, and queue time were included in the models as continuous variables, with a natural log transformation applied to the dwell and gap times. A significance level of 0.10 (α = 0.10) was considered for this analysis when assessing statistical significance.

Results and Discussion

The results for the binary logistic regression model for movement legality of vehicles exiting the McDonald’s driveway with the DAD in place along with one of the five auxiliary test signs are presented in Table 4.

Table 4.

Model Results for Legal Movement Analysis with Logistic Regression

Parameter	Beta	Standard error	Exp(beta)	p-Value
Intercept	1.606	0.454	4.983	<0.001
Sign = Proceed	Baseline
Sign = Turn only	0.009	0.300	1.009	0.975
Sign = Stop and wait	0.178	0.212	1.195	0.402
Sign = WAIT for arrow	0.506	0.231	1.659	0.029
Sign = WAIT turn only	0.455	0.248	1.576	0.066
Queue position = 1	Baseline
Queue position = 2	−0.166	0.226	0.847	0.461
Queue position = 3	−0.139	0.225	0.870	0.537
DAD display = solid	Baseline
DAD display = right	2.021	0.186	7.546	<0.001
DAD display = left	2.305	0.225	10.024	<0.001
Full stop = yes	Baseline
Full stop = no	0.223	0.178	1.250	0.211
Vehicle type = PC	Baseline
Vehicle type = SUV	0.008	0.203	1.008	0.967
Vehicle type = van	0.311	0.308	1.365	0.312
Vehicle type = pickup truck	−0.189	0.228	0.828	0.406
Vehicle type = other	0.225	0.785	1.252	0.774
Platoon = precede	Baseline
Platoon = join	−2.073	0.168	0.126	<0.001
Vehicle direction = left	Baseline
Vehicle direction = right	0.238	0.165	1.269	0.151
Vehicle time in queue	0.004	0.003	1.004	0.120
Ln(dwell time)	0.114	0.061	1.121	0.060
Ln(gap time)	−0.711	0.134	0.491	<0.001
AIC	1109.2

Note: PC = passenger car; AIC = Akaike information criterion; DAD = driveway assistance device; SUV = sports utility vehicle.

The results from this analysis provide further evidence of the influence of the various auxiliary signs on legal movements. As can be seen from Table 4, among the five different signs analyzed, those signs that included “WAIT” messages (i.e., “WAIT for Arrow” and “WAIT Turn Only”) were associated with a statistically significant increase in the likelihood of legal movements compared with the base condition (i.e., the MDOT standard “Proceed” sign). Interestingly, the other two signs including the “Turn Only in Direction of Arrow” sign without the “WAIT” message had only a slight, and not statistically significant, improvement in legal movements. The “WAIT for Arrow” and “WAIT Turn Only” signs increased the likelihood of a legal movement by 65.9% and 57.6%, respectively, compared with the base condition. Anecdotally, the “WAIT Turn Only” sign also showed the highest rate of safe movements out of all sign combinations tested in the field study. Given, overall, the strong positive impacts of the “WAIT” message on legal turning movements at the DAD, it is recommended that the signs including this prominent message be considered for future DAD experimentation.

Moreover, the DAD display on vehicle arrival to the queue had a statistically significant impact on legal movement. As intended with device operation, when a vehicle arrived during a flashing arrow phase, compared with the solid display, they were much more likely to make a legal movement. Overall, this provides evidence that vehicles either accepted arrows quickly, or when seeing an opposite arrow, correctly waited in the queue until their respective arrow appeared. Conversely, vehicles initially arriving on a solid circular red indication may be tempted to enter based on observing the direction of a mainline platoon, which, although illegal, would not necessarily be an unsafe maneuver, assuming that the offending driver joins the platoon. It follows that the drivers who joined a platoon had much lower odds of a legal movement, nearly an 87% reduction in odds compared with vehicles preceding a platoon. However, this result is likely to be an artifact of the signal phasing at this location, which generally experienced the platoons arriving during the solid circular red indication.

Gap time was additionally found to have a negative association with legal movements (i.e., increases in gap time reduced the likelihood of legal movements) and this relationship was statistically significant, although queue and dwell times did not have any influence on legal movements. One explanation for this finding is that vehicles faced with long gap times may be tempted by a clear line-of-sight on the mainline, which might be less likely to result in a legal movement. The implications of the gap time result could be attributed to the importance of ensuring appropriate relative PTCS distances and maintaining clear stopping points at driveways. This also suggests that lengthy clearance intervals provide additional opportunity for illegal driveway exits, especially when the clearance intervals (and the steady circular red indication at the DADs) are active when the platoons pass by the driveways.

Lastly, other variables, including vehicle type (car, SUV, minivan, pickup truck, and other), vehicle’s queue position (1, 2, or 3), vehicle direction (left or right), and whether the vehicle came to a full stop did not have any significant impact on the likelihood of a legal movement.

Conclusions and Recommendations

The DAD is a relatively new device for regulating traffic at low-volume access points in a one-lane, two-way work zone. Given the novelty of the DAD, little is known about its optimal design and operation, particularly related to auxiliary sign messaging strategies. To address this, a field evaluation was conducted to ascertain the auxiliary sign design that best directed motorists to a legal and safe exit from a low-volume access point to the mainline within the work zone. The evaluation included a rotation of five auxiliary sign messages tested at a McDonald’s driveway to evaluate driver compliance when faced with a DAD with flashing red arrows. The following conclusions and recommendations from the field study are presented below.

First, from a high-level standpoint, the DAD generally resulted in a high proportion of safe movements during the field study, with an overall safe movement rate of nearly 93%, which was consistent with the rates experienced in previous evaluations ( 4 , 5 ). Although the overall legal movement rate (62%) was lower than observed in prior studies, this could possibly be attributed to the DADs being employed at commercial driveways and minor side-streets with higher traffic volumes and longer wait times compared with the prior evaluations. Future research should evaluate the optimal volumes for DAD usage.

The primary objective of the field evaluation was to determine how the auxiliary sign message, used in conjunction with the DAD and an NTOR sign, affected legal movement rates for drivers exiting an access point to the mainline. Five different auxiliary signs, each with a different message, were rotated for testing alongside the DAD. The auxiliary sign messages included the current standard sign used by MDOT, along with four other experimental signs, each of which were found to increase the likelihood of a legal movement compared with the standard MDOT sign. The two signs with a prominent “WAIT” message (Figure 5, d and e ) showed the greatest improvements, as the liklihood of a legal movement increased by 65.9% and 57.6%, respectively, over the standard MDOT sign. It is worth noting that the “WAIT Turn Only” sign (Figure 5e) also experienced the highest safe movement rate of all sign combinations. Thus, it was concluded that the inclusion of a prominent “WAIT” message had a positive impact on influencing legal movements at a DAD. Further experimentation with DAD auxiliary signs should continue to evaluate signage that includes the prominent “WAIT” text above the primary message to improve compliance with the steady circular red indication.

Recommendations for Future DAD Implementation and Evaluation

Overall, the results of this field study indicated that the DAD is a device that resulted in a high rate of safe movements, the majority of which were legal movements at low-volume access points along one-lane, two-way work zones. Thus, DADs are recommended for continued experimental use at work zones that include one-lane, two-way traffic, where it is not practical or feasible to provide continuous flagger operation. Furthermore, auxiliary signage should be implemented in conjunction with any implementation of a DAD. This signage may be posted on the DAD pole itself or positioned separately beside the DAD. Such signage should utilize messages that provide direction for what action drivers should take both during flashing arrow phases and steady circular phases. For the flashing arrow phase, sign messaging should include the instruction “Turn” rather than “Yield.” For the steady circular red phase, a separate NTOR sign should be utilized in combination with any auxiliary sign, or as a standalone sign if no other auxiliary sign is provided. The prominent “WAIT” message may be included at the top of the auxiliary sign to further enhance the appropriate action during the steady circular red.

Based on the results of this research, the two auxiliary signs shown in Figure 7 are recommended for implementation alongside future DAD installations. Of these, the “WAIT Turn Only in Direction of Arrow” sign had the highest rate of safe movements of all signs tested in this study. It is also important to note that these signs may be used in conjunction with red or yellow flashing arrows. Although only red arrows were tested in this study, the decision to use red or yellow arrows should be left to the transportation agency, based on the policies and practices within the particular jurisdiction. In many states, the flashing red arrow is either uncommon or unknown, thus leading to motorist confusion, although such arrows are still likely to increase stopping rates. However, at higher-volume locations, it may be more operationally efficient to utilize flashing yellow arrows to eliminate the need for each vehicle to make a complete stop while exiting.

Figure 7.

Recommended DAD auxiliary signage alternatives.

It was evident that controlling the available gap and dwell times at DAD access points would add value in influencing legal, safe movements. Suboptimal timing led to vehicles entering the mainline before the arrival or soon after clearance of the mainline platoon during a steady red indication. Thus, the DAD should be timed and synchronized with the PTCSs in such a way that gap time is optimized to allow platoon clearance, but deters drivers from joining the end of a platoon illegally on the steady circular red. To remedy this situation, two methods are suggested: 1) provide longer arrow phases to allow a greater number of legal and safe movements, or 2) lengthen all-red phases to increase the likelihood that a platoon jumper will make a safe movement.

Providing left- and right-turn storage lanes could improve DAD operations at higher-volume access points. As experienced in this field study, impatient queued motorists often created ad hoc turn lanes if their desired turn was blocked by the queue leader. This led to numerous illegal and/or unsafe movements. At wider access points, and especially those with higher traffic volumes, separate left- and right-turn lanes could reduce the possibility of long queues by improving flow and also reducing unsafe movements.

MUTCD mentions that a “Stop Here on Red” (R10-6[a]) sign is required for an AFAD ( 7 ). Stop locations in this field study varied widely across sites, but even between individual vehicles at the same site. Testing the DAD with an R10-6(a) sign to establish a standard stop location could benefit comprehension and compliance. This may also provide a way to control access point sight lines and reduce the motorist’s ability to judge the mainline platoon during steady red phases, thus discouraging illegal behavior. Lastly, alternate DAD designs, such as using radar detection to signal both mainline directions, could be explored in future research.

Footnotes

Author Contributions

The authors confirm contribution to the paper as follows: study conception and design: T. Gates; data collection; J. Hankin, T. Holpuch, M. Motz; analysis and interpretation of results: M. Chakraborty, J. Hankin; draft manuscript preparation: J. Hankin, M. Chakraborty, T. Gates, P. Savolainen. All authors reviewed the results and approved the final version of the manuscript.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by the Smart Work Zone Deployment Initiative, which is part of the Federal Highway Administration’s Transportation Pooled Fund Program (study no. TPF-5[438]) and is administered by Iowa State University.

ORCID iDs

Jordan Hankin

Meghna Chakraborty

Timothy J. Gates

Peter Savolainen

Travis Holpuch

Matthew Motz

References

National Safety Council. “Motor Vehicle Safety Issues”. https://injuryfacts.nsc.org/motor-vehicle/motor-vehicle-safety-issues/work-zones/. Accessed June 25, 2023.

Ullman

G. L.

Levine

S. Z.

An Evaluation of Portable Traffic Control Signals at Work Zones. Transportation Research Record: Journal of the Transportation Research Board, 1987. 1148: 30–33.

Daniels

Venglar

Picha

Feasibility of Portable Traffic Signals to Replace Flaggers in Maintenance Operations. Report No. TX-00/3926-1. Texas Department of Transportation, Austin, 1999.

Finley

M. D.

Songchitruksa

Sunkari

S. R.

Evaluation of Innovative Devices to Control Traffic Entering from Low-Volume Access Points within a Lane Closure. Report No. FHWA/TX-13/0-6708–1. Texas Department of Transportation, Austin, 2014.

MDOT. Permission to Experiment – 6(09)-35 (E) - Driveway Assistance Device. Michigan Department of Transportation, Lansing, 2019.

Finley

M. D.

Fitzpatrick

Rista

Brewer

M. A.

Avelar

R. E.

Pike

A. M.

Barrette

T. P.

Dutta

Traffic Control Device Analysis, Testing, and Evaluation Program: FY2019 Activities. Report No. FHWA/TX-19/0-6969-R2. Texas Department of Transportation, Austin, 2020.

Manual on Uniform Traffic Control Devices (MUTCD) for Streets and Highways. Federal Highway Administration, U.S. Department of Transportation, Washington, D.C., 2012.

Manual on Uniform Traffic Control Devices, 2011 Michigan MUTCD, 2009 Federal Edition. Michigan Department of Transportation, Lansing, 2013.

Schrock

S. D.

Patil

S. S.

Fitzsimmons

E. J.

Portable Traffic Signals in Conjunction with Pilot Car Operations at Two-Lane, Two-Way Rural Highway Work Zones. Transportation Research Record: Journal of the Transportation Research Board, 2016. 2555: 72–80.

10.

Schrock

S. D.

Patil

S. S.

Fitzsimmons

E. J.

Model Development for Use of Portable Traffic Signals in Conjunction with Pilot Car Operations at Two-Lane, Two-Way Rural Highway Work Zones. Transportation Research Record: Journal of the Transportation Research Board, 2016. 2555: 19–27.

11.

Finley

M. D.

Theiss

Field Evaluation of Pilot Vehicles and Portable Traffic Control Signals With and Without a Flagger. Transportation Research Record: Journal of the Transportation Research Board, 2017. 2617: 1–8.

12.

Evaluation of the Driveway Assistance Device (DAD) Systems in One-Lane Two-Way Work Zone. https://rip.trb.org/view/1741272. Accessed September 29, 2021.

13.

FHWA MUTCD Official Rulings. https://mutcd.fhwa.dot.gov/orsearchresults.asp. Accessed July 19, 2022.

14.

Gates

T. J.

Hankin

Chakraborty

Mahmud

M. S.

Savolainen

P. T.

Holpuch

Motz

Effective Signing Strategies and Signal Displays for Work Zone Driveway Assistance Devices (DADs). Final Report. Smart Work Zone Deployment Initiative, Iowa Department of Transportation, Ames, January 2022.