Click to edit Master text styles
Second level
Third level
Fourth level
Fifth level
‹#›
Today’s agenda
Carl Andersen
Connected Vehicle Program Manager, FHWA Office of Research, Development, and Technology
DMA Program Overview
FRATIS Bundle Overview
Prototype Description and Current Project Status
Robert Rupert
Team Leader, FHWA Office of Operations
Current Project Status of Impact Assessment
Testing Results and Impacts/Benefits from IA
Stakeholder Q&A
We can only answer the questions related to the DMA program.
We cannot answer any questions related to the CV Pilots.
2
The ITS Research Program includes work in three broad areas: applications, technology and policy.  All three of these components must be developed in order to realize a Connected Vehicle Environment. 
The application work can be further described as covering Safety, Mobility, and the Environment.
4
DMA PROGRAM APPROACH TO OVERCOMING TWO KEY CHALLENGES TO APPLICATION DEPLOYMENT
Challenge 1 (Technical Soundness)
Are the DMA bundles technically sound and deployment-ready?
Create a “trail” of systems engineering documents (e.g., ConOps, SyRs)
Share code from open source bundle prototype development
(OSADP website: http://www.itsforge.net/)
Demonstrate bundle prototypes (in isolation)
Field test integrated deployment concepts from across CV programs
Challenge 2 (Transformative Impact)
Are DMA bundle-related benefits big enough to warrant deployment?
Engage stakeholders to set transformative impact measures and goals
Assess whether prototype shows impact when demonstrated
Estimate benefits associated with broader deployment
Utilize analytic testbeds to identify synergistic bundle combinations
5
The six DMA bundles and applications are presented in this slide. Today’s Webinar will focus on the FRATIS bundle of applications.
6
This slide illustrates the location and status of the six DMA bundle prototype development activity.
There are three live FRATIS prototype demonstrations: Los Angels, California, South Florida, and Dallas-Fort Worth, Texas.
7
USDOT sponsored the FRATIS Prototype and Small-Scale Demonstration largely in part because of the major issues in goods movement efficiency, such as
Lack of information sharing between trucking and terminals significantly impedes intermodal freight system efficiency
Lack of freight-specific traveler information such as terminal wait times and dynamic routing options
In August 2012, the USDOT’s Dynamic Mobility Applications (DMA) Program initiated the development of a prototype of the Freight Advanced Traveler Information System (FRATIS) bundle of applications, and a small-scale demonstration of the prototype for assessing the effectiveness and impacts of a regional-based FRATIS implementation. The FRATIS concept seeks innovations to transform freight mobility, including methods to:
•Leverage freight mobility information technologies under development in the private sector regarding freight traveler information, dynamic routing, and load matching;
•Integrate these technologies with public sector Intelligent Transportation Systems (ITS) and sensor information available for roadways in major metropolitan regions; and
•Facilitate accelerated public-private deployment of FRATIS applications.
 
9
Freight-Specific Dynamic Travel Planning and Performance
This application bundle seeks to include all of the traveler information, dynamic routing, and performance monitoring elements that users need.
§Enhances traveler information systems to address specific freight needs
§Provides route guidance to freight facilities, incident alerts, road closures, work zones, routing restrictions (hazmat, oversize/overweight), and performance monitoring
§Builds on the Cross-Town Improvement Project (C-TIP) Real Time Traffic Monitoring (RTTM) and Dynamic Route Guidance (DRG) applications for best route between freight facilities.
§Provides intermodal connection information, container disposition and schedule
§Leverages existing data in the public domain, as well as emerging private sector applications to provide benefits to both sectors.
§
Drayage Optimization
This application bundle seeks to combine container load matching and freight information exchange systems to fully optimize drayage operations, thereby minimizing bobtails/dry runs and wasted miles, as well as spreading out truck arrivals at intermodal terminals throughout the day.
§Reduces freight delays at key facilities that overbook their capacity to ensure uninterrupted operations within the terminal/warehouse
§Optimize drayage operations so that load movements are coordinated between freight facilities
§Individual trucks are assigned time windows within which they will be expected to arrive at a pickup or drop-off location
§Early or late arrivals to the facility are dynamically balanced
§Web-based forum for load matching provided to reduce empty moves
10
I’m going to walk you through the six core FRATIS functions in order to give you a better overview of the whole process.
11
The first step starts just the same as if FRATIS was not in place with the drayage companies entering their daily pick up and drop off orders into their back office systems (if it exists), if they don’t have a backend system, they can use their own FRATIS user interface to enter their daily orders.
12
Those orders are collected, sent to the FRATIS optimization algorithm.  Various methods were used to move this data between the drayage companies and the Optimization algorithm including entering the data directly into FRATIS and FTPing the information in files.   Once the information was gathered the optimization algorithm was run for a single day and a single drayage company.  The algorithm optimized to reduce the number of bob-trail trips and the distance trucks would have to travel during the day while also taking in to account constraints by the drayage company like driver availability and hours of service, time windows, etc.
13
Once the optimization is complete the drayage company can review the optimized plan.  This is a important step as it allows the drayage company time to validate the plan and make any changes to it before they send it out.
14
The approved plan is then broken up and each driver’s part is either 1) sent to the truck through in vehicle units which were TomTom 7150 units (Los Angeles) or smartphones  (Dallas-Fort Worth)  or 2) manually assigned with legacy system (South Florida) for the FRATIS prototype.  This information included each of the drivers pickup and drop off points for the day in order with their routes for each job.  Routing information would update dynamically as needed should traffic conditions change to make sure drivers where taking the fastest route. 
15
Any updates to the orders would be sent directly to the drivers units so they can see the changes. The in vehicle units also allowed for direct communication between the drivers and dispatchers which was used to confirm changes and ask any questions about orders.
16
As orders are completed the drivers update their information on the in-vehicle unit giving status updates to the drayage companies about their progress.  When customer contact the drayage company about their order status the company can quickly review the truck progress and give a detail ETA for the order based on the trucks location and status.  This feature can be enhanced potentially in future versions of FRATIS where ETA information can be directly sent to the customer based on the trucks status. Another feature for future enhancements includes optimizing on each driver within a fleet.
Once all of the orders are filled the same process starts over again the next day.
17
USDOT selected  multiple locations to test the FRATIS prototype under different environments that incorporated innovative, unique features of the FRATIS bundle
Development teams:
LA – Cambridge Systematics
DFW – Leidos
SFL – Cambridge Systematics Team
Under separate contract USDOT contracted with Productivity Apex, Inc. (PAI) to develop the Optimization Algorithm, which is used by all three sites.
LA and DFW were the first to get started; followed a short time later by SFL. FRATIS prototype development began in August 2012, baseline data collection began in the August-September 2013 timeframe, and prototype demonstration for LA and DFW began in early 2014. LA ends in Feb 2015 and the other two in spring 2015.
In LA, the prototype will be implemented around marine terminals to move cargo out of the port more efficiently; used Bluetooth (wifi)- based terminal queue management system. In Dallas-Fort Worth, drayage operations will be coordinated among rail and local truck drayage companies. Also, Bluetooth and DSRC technologies used for terminal queue time measurement. In South Florida, in addition to the two FRATIS applications, use of freight transportation for emergency preparedness and response efficiency will also be examined, including use of targeted real-time information for natural disasters, including pre-event staging of supplies, post-event relief delivery coordination, and critical road/facility closures.
All products developed under this project will be made available as open source through the DMA Program’s OSADP or the DCM Program’s RDE.
 
18
19
The LA gateway region project consisted of two key partners: Port Logistics Group and Yusen Terminal Inc  at the Port of LA, and  several other stakeholders including LA Metro, Gateway Cities COG, and Harbor Trucking Association.
Port Logistics Group –is a drayage and logistics firm, with a fleet of 50 vehicles moves 25,0000 TEUs per year,
50 trucks installed with TomTom Link 510 and TomTom 7150 GPS Truck
Yusen Terminal – Major intermodal terminal at Port of LA handles 1,400 containers per week, testing new information exchange with Port Logistics Group, as well as allowing queue measurement sensors  (Bluetooth) to be placed at their terminal approach (and inside)
FRATIS LA Test
Enabled by a unique regional public-public partnership – the Gateway Cities ITS Working Group – that has develop and overall freight ITS and connected vehicle program plan for the region
Facilitated by LA METRO, the Gateway Cities COG and the   Harbor Trucking Association
Designed based on extensive user feedback from dispatchers, drivers and marine terminal operators
Deployed and operated successfully since early 2014, with continuous system enhancements and expanded use over time.
An example to the nation of how to successful plan, design, deploy and test advanced ITS and connected vehicle technologies
LA Drayage Optimization and Freight-Tailored Traveler Information
Daily optimized schedules per driver based on average stop times, predicted travel times, expected terminal wait times, and other constraints
Real time terminal queue info, driver messaging, and traffic; dynamic routing for trucks through in-cab navigation TomTom devices
20
The unique feature of the LA Gateway Region FRATIS prototype was the interface between the drayage company (Port Logistics Group) and the  marine terminal operator (Yusen Terminal).
The interface development created a two-way messaging between terminal and drayage firm with estimated time of arrival (ETA) for dray approaches and MTO-dispatcher messaging and alerts.
Benefits of FRATIS Trucking – MTO Communications System Testing
If deployed on a large scale, and supported by all parties (including shippers), has the potential to radically improved port terminal and trucking efficiencies
Through “dynamic appointments”
Has successfully brought together the trucking and terminal operations communities in the ports region
A major positive development
21
In Dallas-Fort Worth, drayage operations are coordinated among rail and local truck drayage companies.
The DFW project has four major stakeholder participants
IMCG (container yard):
Install of wait time hardware
Provider of equipment status
Receiver of terminal notice from drays
Associated Carriers (drayage company):
50 trucks in DFW fleet
40 equipped with TomTom Link 510 devices for prototype monitoring
Southwest Freight International (drayage company):
~50 trucks in DFW fleet (46 in 2012 at pilot start)
10 equipped with TomTom Link 510 devices for prototype monitoring
BNSF:
Receiver of terminal notice from drays
Providers/Vendors
Trinium (software):
Dispatch software used by both Associated and Southwest
Web app MC2 which provides routing, nav, traffic and weather
Acyclica:
Bluetooth/Wi-fi wait time hardware and software
DFW prototype deployed in January 2014 and small scale demonstration is active.
22
DFW used Bluetooth (wifi) and DSRC technologies to calculate terminal queue times. This was the only site to use DSRC technology.
Partners
USDOT ITS Joint Program Office (Connected Vehicle Test Bed)
IMCG (facility and dray company) Trucks were in and out of the yard frequently.
Process:
Install equipment (stationary roadside unit and DSRC radios on trucks)
Configure connected vehicle data management system to store/query data
Develop code to calculate relevant metrics:
Wait time
Time on yard – active vs. idle
Comparison to wi-fi wait times
DSRC equipment installed for 30-day pilot.
Data collected being analyzed, results are not yet available.
23
In South Florida, in addition to the two FRATIS applications, use of freight transportation for emergency preparedness and response efficiency will also be examined, including use of targeted real-time information for natural disasters, including pre-event staging of supplies, post-event relief delivery coordination, and critical road/facility closures.
The primary partner for the SFL test is Florida East Coast (FEC) Highway Services – the drayage arm of  FEC Railway.
Overview of FEC Highway Services
Fleet of 100+ trucks
Operations in Miami, Ft. Lauderdale, Jacksonville, Atlanta, and others
Approx. 63 in trucks in Miami
Uses Qualcomm for fleet management
Includes 2-way messaging capability
~90K moves per year, mostly as part of a larger rail move
Uses customer appointments to schedule loads when a train arrives
How was drayage optimization tested?
Loads transmitted daily from FEC system to drayage optimization tool via secure FTP server
FEC dispatch selects the appropriate loads and drivers within the web-based interface and runs the drayage optimization
FEC dispatch uses results to assign loads manually with legacy system
Technologies used for FRATIS South Florida:
Drayage Optimization (integrated load matching and freight information exchange that helps maximize efficiency of daily drayage work plan)
50 TomTom devices installed on partner  trucks with WebFleet subscription
Web-based drayage optimization tool
Data mapping look up table linking partner  data to optimization tool
Dedicated, encrypted FTP server
For Emergency Preparedness and Response Efficiency the technologies used were
Google Map-based emergency management software and server
Google Map-based emergency management Android application
Android smart phones/devices
24
Unique feature - Increasing Emergency Preparedness and Response Efficiency App
•Three test scenarios were developed representing progressively worsening hurricane conditions
•As the scenarios increase in severity, the conditions reported by users were anticipated to increase in severity
•Each different user type (emergency management, truck drivers, private business/freight hubs) had different
reporting responsibilities
•Each scenario was completed over the course of one day
•Participants used an Android device or internet browser to complete condition reports, comment on other condition
reports, and update business/terminal status
25
Institutional issues, as is the case with most projects, tend to be more challenging than the technical ones.
Institutional Challenges
1. Strong partnership and continuous commitment to the project. Engaging Key Partners…and Keeping Them Engaged is Critical!
Engagement requires:
Regular communication
Minimal impact to operations
Proof of benefit
2. Operational disruptions and partner staffing availability. Partners volunteering to participate in the demonstration are focused on day-to-day operations and often have little time to devote to new technologies/new procedures/training; staff cannot devote full time to the test. External factors that impact not just the  prototype testing, but with daily operations at the terminal, drays (e.g., LA PierPass (fee required for cargo movement during peak hours), unions, lack of equipment (chassis shortage in LA area recently)
3. Driver and dispatcher acceptance – drivers often think “Big Brother is Watching” with new devices installed, GPS tracking. Dispatcher trust method and procedure they have been doing to process and assign orders, and often lack trust in plans generated the optimization. There is hesitation to use the new system.
Technical challenges
1.Many drayage companies had different back-end systems which made it impossible to have a single method to pull order information into the optimization algorithm.  This required custom methods to be created for each drayage company which sometimes required data to be entered twice once into the drayage company order system and second time into the FRATIS system.
2.Customization of optimization program for each site. Each dray company operates differently. The optimization program can be customized to handled specific constraints such as driver preferences, high priority customer needs, trip types; however, this required time and funding resources.
3.LA and DFW prototype development teams experienced  Bluetooth equipment outages. Outages were due to 1) weather: protective enclosures  were provided, but when power was lost at a site, readers were still impacted; and 2) staff interference (unplugging).
These are from the prototype development team perspective. As you will hear from Bob shortly, the impacts assessment team independently-generated similar lessons learned.
26
Documentation from FRATIS phase 1 such as the Concept of Operations, system requirements. research assessment, and test readiness assessment are available. For someone unfamiliar with FRATIS, the best place to start is with the ConOps.
For Phase 2, the Architecture and Implementation Report  and the Demonstration Plan for the three FRATIS prototypes small-scale demonstrations are available from the Connected Vehicles Pilot Deployment Project (link provided on the slide).
A final report will be developed for each of the three prototype  projects at the conclusion of the demonstrations.
Code for the drayage optimization algorithm, and source code for the three prototypes is available on the Open Source Application Development Portal (link provided on the slide).
27
The FRATIS Impact Assessment contract was awarded at the same time as the 3 prototype contracts.  The impacts assessment team consists of CDM Smith, Booz Allen, and North River Consulting.
The interaction between the assessment team and each of the development contractors was useful to both parties.  Because of start-up issues, the South Florida prototype received less attention during the familiarization period than did DFW and LA.   PAI’s development of Excel modules for use by the assessment team was very helpful.
The purpose of the impact assessment was to measure reduction in the series of FRATIS performance measures and transformative targets previously identified in previous freight projects and FRATIS phase 1.
29
This shows three time periods of assessment team activity.  The first was familiarization with the 3 prototypes including attendance at kick off meetings at each site.  From these interactions, the assessment team prepared an assessment plan that was approved during 2013.  Also during this period, the assessment team worked with PAI on Excel-based automated tools that would analyze the Tom-Tom being collected at the sites.
After the assessment plan was completed, there was a relatively quiet period for the assessment team while development progressed.  At DOT request, some preliminary analysis related to the LA prototype was conducted first in the summer 2014 and then in the fall.  The assessment team analyzed the LA dray company fleet trips for each day from Labor Day 2014 through Thanksgiving 2014. During the 82 days, data for 18,542 trips was collected.
Despite efforts by the impact assessment and prototype development teams, most of the impact assessment remains to be completed.  The assessment contract was extended to August 2015.
The two key parts of the assessment at each site will be analysis of data collected by the development contractors and interviews with the dray companies.  Based on the interviews and with information about the dray company population in each area, the assessment team will estimate regional impacts. 
30
These are the principal hypotheses for the FRATIS assessment as derived from the approved Assessment Plan.  The first several deal with actual operational improvement as shown in the test data.  Some of these will be supported by insights gained in the interviews with the stakeholders.  The last two hypotheses deal with the regional expansion, but it will take future efforts to really determine whether or not additional companies actually implement FRATIS.  The FRATIS Assessment Plan contains more detailed breakout of the hypotheses and how they will be analyzed.
31
The two principal assessment methods are to analyze the test data collected at each site from Tom-Tom devices purchased by the FRATIS program, and interviews with stakeholders. 
There are two primary types of test data being collected at each site.  The GPS-based Tom-Tom device collects continuous movement and location data that is used for calculating travel time between point for each truck in the fleet equipped with a Tom-Tom device.  Each site also installed devices at various points in intermodal and  port terminals so that queue time data can be calculated.
The interviews will be with drayage company dispatchers and drivers as well as terminal operators and public sector stakeholders at each of the three prototype sites.  Taken together with the test versus baseline data, this will provide the information needed for the assessment. 
32
The intent of the assessment is to compare “apples to apples” in the baseline period before FRATIS and after FRATIS is being used by the dray company.  Early in the analysis, the assessment team, in conjunction with Noblis, identified several operational conditions for which data was collected both during baseline and test periods.  The principal operational conditions to be examined are traffic conditions on the road segments where the trucks operate and the volume of business the drayage company was conducting as measured by daily orders for container movements.  The assessment team did some preliminary bin analysis, but found no appreciable differences.  DOT thus told the team to concentrate on trip times.
In order to confirm the integrity of the assessment and to test the realism of the data analysis tools, the assessment team did some preliminary analysis of several months of baseline and test data for LA.  DOT also asked that the assessment team prepare weekly reports of trip time analysis during the fall to facilitate testing underway at the dray company.  The next two slides contain some of the information from those weekly analyses for LA. 
33
This slide is an extract from the analysis tool that the assessment team uses to determine trip times.  On a typical day, there are 35-45 drivers who have trips recorded with the TomTom devices.  This extract shows the first four trucks on the list.  Data reported by the tool includes the total miles driven by the driver in the day, the amount of time involved in the trips, and averages.  The data includes individual trips by origin and destination.
One preliminary analysis the assessment team has done involves segmenting the data by length of trip.  While many trips are under 10 miles, there are some trips in excess of 75 miles.  The team created four distance bins and may find in the later analysis that comparisons between like distances may be meaningful.
It is also interesting to note that there was no noticeable effect on trip time from the congestion and heightened port activity in LA-Long Beach during that period.  The trip ends when enters a queue, so terminal delay itself is not part of the trip time calculated from the TomTom data.  Nevertheless, it seems surprising that the road trip itself was unaffected by the operating conditions.  This will be examined in more detail during the impact analysis.
34
This chart does compare baseline days with days during the test period.   The overall improvement between the two periods was 3.73%.  These results are preliminary in that the assessment team has not discussed the details with the stakeholders nor determined in any definitive way why the average trip time decreased.  It may be that drivers were more careful during the test period, but it should be noted that during the June-November 2014 period the optimization algorithm was not in use and there were various problems with the communications between the dispatchers and drivers and between the dray company and the terminal.  Thus, FRATIS use is not the primary cause of the lower trip time.
The assessment team’s effort going forward will be more detailed analysis of WHY the trip time changed.   If the team finds that the 3.73% reduction holds and is attributable to FRATIS, then the team will determine the economic value of that benefit to report with the final results.
At this point, the chart shows how baseline versus test data can be compared and displayed.
35
Once the prototype test has been completed at LA and DFW, the assessment team will fully analyze the daily test versus baseline comparison data and produce quantified and non-quantified benefits and impacts of FRATIS.  If resources permit, the assessment team will also do comparative analysis of South Florida data.  Armed with those impacts, the assessment team will investigate and estimate what the impact might be if FRATIS is expanded to other users in the region.
The expansion could be additional customers and partners to test participants or could be additional drayage companies or additional terminals within the port complex.  The assessment team hopes to be able to expand the impact assessment to other ports and regions to try to quantify what it might mean to expand deployments of FRATIS.  This will depend on the availability of information in each region as well as resources with the contractors on the assessment team.
36
When the impact assessment is complete,  there will undoubtedly be many lessons learned to report.  The five preliminary lessons learned show in this slide are based on assessment team observations in the planning and execution phase of the testing.  We hope these preliminary lessons learned are valuable to the DMA program as it begins to define additional prototypes and connected vehicle tests.
What should be clear from these preliminary lessons learned is that it is hard work to overlay a prototype system on day to day transportation operations.  It is worth noting that a key complaint about early FRATIS in LA was that because it was not integrated with existing systems duplicate data entry was required.  A late fix to the system provided the needed integration.  Although it is more expensive and probably takes longer to integrate the old and new capabilities, the likelihood of success in the test increases dramatically.
Devices like the TomTom 510 are particularly useful in prototype test impact assessment because they provide accurate data without anyone having to taken any action.  Analysis tools make it easier to process the data, but such automated approaches to data collection are definitely the way to go.
Working in an operational environment has risks that DOT and its contractors must be willing to live with.  In LA, the on going congestion and chassis problems at the ports adversely affected participation.
37
Although the assessment team will have much more to say about expansion potential, the need to integrate with existing systems, as mentioned earlier, is essential. 
It is also clear that individual companies by themselves cannot fully implement the kinds of changes that a FRATIS-like system would involve.  Sharing of information and making management decisions that affect other companies or that seek to optimize or improve overall port operations may have greater public or regional benefits than individual company benefits.  To really implement a FRATIS-like system widely is very likely to require a central authority to serve as “honest broker” in making a port or regional-wide system work.  This could be a company created for the purpose, such as Pier Pass is in LA, or it could be something  a port authority or metropolitan transportation organization could do itself.  The assessment team will explore this organizational issue in more detail in its impact assessment.
38