Progress Update

2008 October

Instrumentation/Analysis • Empty Containers • Stack Management • Data Support • Commercialization

The mission of the MeTrIS project is to develop geospatial tracking and modeling technologies to smooth the flow of freight through the San Pedro ports and the Los Angeles basin. Our work does not end with GPS tracking; it begins with GPS tracking.

The project is funded by the USDOT and cost share from the investigative team and partners.

Instrumentation and Analysis

We have been following a small fleet of drayage trucks serving the San Pedro ports. The first truck was instrumented with an onboard data logger in August 2007. In November 2007 we began testing custom-manufactured real-time reporting units. We have acquired 100 units and are in the process of installing them. We will ramp up to 500 trackers in the coming months.

Our units sample truck locations at intervals of about 12 seconds, creating detailed traces of routes, stops, speeds, etc (by comparison, commercial tracking is typically at 5-15 minute intervals). To date we have more than 130 truck months of data and have developed preliminary analytical tools. Several original GPS-enabled information products have been created, including:

• Terminal turn times, measured not only from the moment of admission at the terminal gates, but also from when trucks join the queue outside a terminal
  
• Traffic "drainage:" proportion of drayage traffic exiting along a freeway or thoroughfare
• Speed on freeways and other port access routes, and speed variability, by time of day
  
• Detour behavior when traditional routes are closed or congested
  
• Use of non-truck-designated streets

A web page of sample outputs is publicly accessible.

Lead: Digital Geographic Research Corporation

Model: Empty Container Management

Deadhead trips (bobtails, bare chassis and empty container hauls) account for roughly half the traffic on port access routes such as I-710. Our Optimal Storage of Empties Model (OSEM), developed by Dr Richard Church and Ting Lei at UC Santa Barbara, shows that 25% of deadhead trips can be eliminated, by establishing strategically placed depots for empty containers. There are practical barriers that may limit this degree of benefit, including container branding, per diem charges, and motor carriers' revenue models. The magnitude of common potential benefit, in efficiency as well as air quality, may help to overcome these policy barriers.

Model variants consider various possible management practices, such as hauling containers to the nearest depot (trucker-optimal) with a separate activity to balance depot inventory, versus a system-optimal strategy that would involve an authority that directs trucks to designated depots and dynamically advises consignees and shippers of potential street turns. Future work will develop decision support models to permit multiple stakeholders to examine the individual and system-wide costs and benefits of each strategy, and depot location scenarios.

Lead: University of California, Santa Barbara

Model: Container Stack Management

Much of the delay at port and rail terminals arises when containers have to be retrieved from grounded stacks. Murphy's Law holds that the container requested is invariably at the bottom of the stack, requiring upper boxes to be repositioned. This rehandling forms a large proportion of  crane moves, the exact numbers depending on stack height.

With knowledge of estimated time of arrival (ETA) of trucks, our container stack management models, developed by Dr Anne Goodchild and Wenjuan Zhou at the University of Washington, show that rehandles can be reduced by 25-50%, with considerable ramifications for reducing truck wait/idle time, and improving terminal velocity and air quality. Generally speaking, containers expected to be picked up soon are positioned near the top of the stack. Again, there are barriers, from labor implications to truck queue management. And again, the variety and magnitude of benefits will be a powerful argument towards overcoming the barriers.

Lead: University of Washington

Data Support

Dr Mike Goodchild and Linna Li at UC Santa Barbara have developed a data model for geospatial representation of port drayage. A data model specifies the data elements of essential entities (e.g. ports, terminals, trucks, containers, chassis, warehouses), their essential attribute fields and their relationships to one another, facilitating data handling and exchange. The ESRI UNETRANS model, a generic template for transportation data, was developed 6-8 years ago by the UC Santa Barbara research team. The drayage data model is a vertical UNETRANS-compatible component.

Our conflation solution matches one street database to another, blending the positional advantages of one with the attribute accuracy of the other.

There is always positional uncertainty in geospatial data. The locations of roads and other facilities (like most other data, e.g. population, corporate earnings) are specified with some degree of uncertainty, that is tolerable for some applications but perhaps not for others. Road maps are typically up to 5 meters (1½ lanes) off their true locations in major cities. When marrying GPS data with infrastructure maps, for example, these errors propagate and cause analytical and modeling problems.

Conflation is related to map matching, the problem of determining the street that a GPS position is on. Due to errors in GPS as well as the map base, map matching can return a position on the wrong block the wrong highway, or the service road rather than the highway. Resolving these errors is essential to our work, and of universal utility to geospatial analysis.

Outreach and Commercialization

All of the above efforts are being developed with the eventual objective of commercial deployment. With our partners, Bill Lyte of CALMITSAC and John Glanville at Athenaeum Capital Partners, we are actively publicizing our research results and developing paths towards commercialization in conjunction with other goods movement technologies and local development initiatives. We welcome opportunities to broaden the circle of partners researching, developing and supporting MeTrIS.

Leads: Digital Geographic Research Corporation with Athenaeum Capital Partners, CALMITSAC

© Digital Geographic Research Corporation	Contact us