Are we there yet?

Imagine a paper road map unfolded on your desk. Then imagine being able to layer any number of transparent maps on top of that, with each one adding new specialized information to the main map below. Smart routing works a little like that, except that all the maps are digital, those additional data layers provide both static and real-time information, and some routing systems will do the map reading and decision-making for you too, even on the fly.

Today mapping and routing has become the business of scientists and computer technicians as well as cartographers. It is an almost unbelievably complex and difficult process, but the results are making it easier for carriers and others to take transportation planning and navigation to new levels of efficiency.

“Accuracy is critical when it comes to routing and mapping,” says Luke Wachtel, business development manager for Maptuit (www.maptuit.com). “However, to be this accurate, especially for trucks, is extremely difficult, challenging and algorithmically complex.”

“Mapping and routing has become a complex science,” agrees Marc Prioleau, vp marketing for deCarta, Inc. (www.deCarta.com). “Even five years ago, routing and navigation was all about you and a map. Now there are so many interesting kinds of data out there, such as traffic data; toll data; information about sharp turns, angles and grades; weather information; points of interest information and so on that people want built into routing systems so that they can do what they need to do. The available data that can be integrated, layer after layer, into routing systems just keeps getting richer all the time.”

THE MAP

A road map, in this case a digital map, still resides at the heart of a routing system, no matter how sophisticated and complex. There are not very many digital map providers in the world and even fewer who provide maps that take into account the unique attributes of trucks and trucking. For digital maps, think NAVTEQ, Tele Atlas, ALK, Rand McNally, Navigation Technologies, Google, GDT and Tiger for instance, and you are close to a comprehensive list for North America, including digital maps with and without truck-specific data.

ALK Technologies, Inc. (www.alk.com) has been providing transportation map data for more than 20 years. In fact, the company released the 21st annual upgrade of its routing, mileage and mapping software solution in May of this year, called PC*Miler 21. According to ALK, the latest version features “significantly improved map graphics and navigation functionality, as well as important database updates.” “Version 21 is a digital map with an extremely comprehensive set of truck attributes,” explains Ed Siciliano, vp sales and marketing for ALK. “Some 21,000 companies use our product today and they've become a tremendous feedback group to help us create a pristine digital map of over seven million miles of highway. Over time, we've adjusted our map and algorithms to take into account various factors. For example, thanks to vehicle tracking, we can see ways our customers actually prefer to go. We have permanently adjusted our map and algorithms to incorporate those preferences.”

Tele Atlas (www.teleatlas.com) also provides digital maps and dynamic content that is the foundation for a wide range of routing and navigation systems. To get the mapping job done, the company relies on a network of some 2,400 full-time staff and contract cartographers at offices in 24 countries along with a network of professional drivers, mobile mapping vans and more than 50,000 other data resources.

“We try to make our map files easy to read and integrate,” says Dana Fenner, director of fleet and logistics for Tele Atlas. “Our mapping techniques and the size of our field force help us to maintain current and correct maps to avoid routing errors. Most people are surprised to learn how many changes take place to the U.S. highway system every year. Two-thirds of the exit numbers in Florida, for example, have been recently changed and renumbered.”

RICH DATA & ROUTING SOFTWARE

The Integration of rich layers of historical and real-time data with digital maps is what really enables the transformation of routing systems into smart routing systems, however. Some digital maps providers, such as ALK, also do their own data integration and routing software development, while other companies integrate maps, data feeds and applications from various providers to create multi-dimensional route planning and navigation systems. Maptuit, for instance, begins with digital maps from NAVTEQ (www.navteq.com), adds their own truck attributes data and then hosts a layer of routing and map updates on top of that, according to Luke Wachtel. “We began with NAVTEQ maps with our truck attributes,” he explains. “Then we added updates received from our customers.

“Several years ago, we were proud of reducing accidents for our customers by 25%. Then one of our fleets said, ‘Have you ever considered changing routing to the easiest route?’” Wachtel recalls. “We made some tweaks to the algorithm to add simplicity of travel as a factor. Now, the system will select a slightly longer route, for example, if it means staying on streets that only change names at intersections, or avoiding roads with forks and oblique intersections — things that make it more likely a driver will get lost or in trouble. After we implemented those changes, our accident reduction improved to about 35%, which is our average now.

“Today every road segment in our system has a number of attributes attached to it, including factors such as tolls, accident data, sharp angles and grades, historical traffic indices, historical weather information and fuel price data,” he adds, “and all that data is constantly changing. The attributes can also be turned on or off depending upon a customer's needs and can be integrated, of course, with a customer's own operating parameters. This information is combined to create a value or cost for a given route.

“These different data flows can constantly change the value of a particular route,” Wachtel continues, “so that four people running the same route at the exact same second might each have a different value for that route. For example, a rookie driver might be routed to stay on highways that permit him or her to avoid difficult turns while a professional, seasoned driver who is very cost-sensitive might want a route that minimizes fuel consumption. Drivers with pressure-sensitive cargo, such as bags of potato chips (and this is from a true incident), might assign a very high ‘cost’ to routes that would take them above a particular elevation, to avoid having the bags of potato chips explode.”

TRAFFIC

According to Maptuit and other smart routing system providers, the addition of real-time or “dynamic” data streams is the next frontier for routing and navigation systems. “Within the past year traffic and other dynamic content has improved dramatically,” says Dana Fenner of Tele Atlas. “There is much more information available now and much more interest in using it. When you combine real-time traffic incident data with historical, predictive data, for instance, that can create a very powerful and useful tool.”

“There are really four types of traffic data,” explains Marc Prioleau of deCarta, “incident data, real-time traffic flow data, predictive or historic traffic data, and dynamic predictive data, which includes information such as event and school schedules.

No matter how much data you have about traffic, the real question, however, is ‘based on all this information about traffic, how should I drive today?

“We've added a system called Traffic Manager onto our platform. It takes all the traffic information we have into account in building a recommended route,” he offers. “Most routing algorithms have a posted speed attribute, but Traffic Manager overlays that. Essentially, it says, ‘the posted speed is this, but ignore it and use this speed in your route calculation now.’”

Incident data (about accidents, stalls, road closures and such) has been around for a relatively long time and video cameras along freeways plus the ubiquitous “traffic 'copters” over major metropolitan areas have improved the timeliness of the information. Historical travel speed or traffic flow information is also relatively easy to come by. It is capturing real-time information about traffic flow, along even secondary roads, that has been the toughest problem to solve.

“With historical traffic information, you can do predictive ETAs,” says ALK's Ed Siciliano. “For instance, we can look at historical driving times over a particular route in the morning, afternoon, during evening rush hour, overnight and on the weekend and do optimized routing based upon this information. We expect to incorporate that type of information into PC*Miler and our CoPilot turn-by-turn navigation solutions. Collecting and processing real-time information about traffic flow along a particular road in time to be of use to an individual driver is a more challenging matter.”

For ALK, the solution may be to use existing in-vehicle devices equipped with GPS as “probes” to capture information about traffic flow that can be shared with other drivers in time to help them make route changes, if appropriate. “Imagine if we could ping GPS devices in vehicles ten miles ahead of you and then present back to you a recommended detour route if need be,” Siciliano says. “That kind of information would be very, very valuable.”

In the spring of 2005, ALK participated in an experiment with researchers at Rensselaer Polytechnic Institute's Center for Infrastructure and Transportation Studies to test the viability of using GPS-based traffic probes. The Center gave 200 volunteers hand-held computers with integrated GPS and Sprint PCS communications. The volunteers kept them in their cars as they drove to and from work or school in the Troy, NY area.

According to ALK, each computer ran CoPilot navigation software, which stored the default route a driver usually took. Every thirty seconds, the unit transmitted its location over the Sprint network to a server, where ALK software computed its speed. If a sudden drop in speed indicated a traffic jam, it transmitted that information to the software in the hand-held computer, which then recalculated the route to avoid the congestion and gave the driver audible directions for a new path.

ALK hopes to commercialize the dynamic route guidance system, notes Siciliano, and its CoPilot software already has the ability to handle it. “There is plenty of data available now,” he says, “but in our opinion it is just not comprehensive enough. We need more probes, more data points to make it a truly viable solution.”

People at Atlanta-based AirSage (www.airsage.com) think they have found the secret to adding more traffic “probes.” In fact, they're so sure that the company has patented their technology, which converts ordinary mobile phone signals into traffic flow information. “We take the network signaling data from cell phones, aggregate it, and match it with computerized street maps via proprietary algorithms to calculate travel times and speeds along specific road segments,” explains Tom Bouwer, vp-sales & marketing for AirSage. “It is important to note that all the data is aggregated and anonymous. We have multiple layers of privacy protection to ensure that no proprietary, customer-identifying data is accessed or released from the secure environment of the wireless carrier.”

Today AirSage provides their traffic data in an XML stream to an application, a navigation device, or to a radio or television station. “We are the only company that has really been able to make this work,” says Bouwer, “and that has just been in the past six months.” (See news story on page 76 for more information.)

POINT-TO-POINT

Providers of navigation technology designed for the general consumer have recently been adding so-called granularity to their systems for some time, including things such as the ability to find restaurants or theatres nearby or even to see which retail outlets within range carry a particular product. For truck fleets, access to good point-to-point information can be even more valuable, and it is increasingly available in routing systems in a variety of forms.

ALK, for instance, provides point-to-point information via a “save location” function, according to Ed Siciliano. “We provide dock-to-dock level detail now,” he says. “For example, if you deliver to a facility that has multiple buildings with 400 different delivery doors, drivers will want to know exactly which door they should go to at that general street address. At the site, a driver can hit ‘save location’ and that exact location information will be stored in the company's routing system. Some companies are also willing to share that information with us, in which case we add it to our own system to share with other users.”

DeCarte plans to launch its own take on point-to-point routing this summer. “We will launch a predictive point-to-point routing solution in early summer,” says Marc Prioleau. “It will be able to answer questions such as, ‘Where are parts stores I can drive to within twenty minutes of here at rush hour?’

“There is a lot of interest in point-to-point information today,” adds Dana Fenner of Tele Atlas. “There are 43-million point addresses in the U.S. today, but there are always gaps and problems. Developers of routing systems are constantly trying to design around those problems. A lot of good information and analysis is coming out of the supply chain systems.”

If knowing exactly which door to knock on is critical to drivers, knowing if roads will be passable along the way is equally important. So it is no wonder that historic and real-time weather information is finding its way into smart routing systems, too.

“Weather is another exciting layer of information,” offers Luke Wachtel of Maptuit. “You could integrate real-time weather information just like real-time traffic information. We haven't added it yet, but we will. Imagine being able to look at a radar weather map over the top of your routing maps or being able to automatically take historic and real-time weather into consideration as you plan routes. Suppose you're trying to decide whether to route a driver down I-70 or I-80. Maybe you typically use I-80, but the system takes into account strong headwinds on I-80 and recommends I-70 instead because of the significant fuel savings.

“Real-time weather information could also be used to tell a driver running along a route where it has recently rained that the temperature is dropping so there might be freezing conditions soon,” he adds. “Our own routing system would automatically take that information into account and downgrade the ‘value’ of that route as compared to others because of the possibility of freezing conditions.”

BUSINESS-SPECIFIC INFORMATION

No matter how much data you have about roads, traffic, weather and other conditions along the way, it all has to be somehow integrated with your own fleet's operations-specific circumstances for smart routing systems to deliver the maximum benefit. At UPS Logistics, for example, there are about 140 data options that can be turned on or off during route planning, according to Cyndi Brandt, technologies marketing manager for UPS Logistics (www.upslogis tics.com).

“We begin with the Tele Atlas map and then use it with our own routing and dispatch software,” she explains. “We do route optimization day by day and weekly, as well. Our typical user will do a download of orders about 3:00 p.m. to 4:30 p.m., create routes for the following day and then send the information to the warehouse so that orders can be picked and loaded and ready to go out the next morning.

“Our system looks at time and distance, plus a lot of other business-specific information, such as the customers' hours of business operation, driver preferences, delivery time windows, number of resources available that day, the order in which you want to use those resources, and so on,” she adds. “Today we have 4,100 installations in 52 countries and we've been a user of Tele Atlas maps for about twenty years now. Our software couldn't do what it does for our customers without them.”

For many fleets today, smart routing systems have become almost as central to daily business as trucks and the drivers who operate them. For others, however, the considerable benefits of implementing a routing system are yet to be discovered. “I am still amazed at the number of companies that don't yet use a routing system,” says Dana Fenner. “There are millions of drivers out there that are still just going from Point A to Point B using paper maps and their knowledge of the area.”

“The ability to do better strategic planning is the great gift of this technology,” observes Chuck Papa, vp information technology for Penske Logistics, a wholly owned subsidiary of Penske Truck Leasing, a joint venture of Penske Corp. and General Electric (www.penskelogistics.com). “For fleets that have not yet made the investment in routing technology, the good news is that it is easier to get started than ever before.”

Routing versus navigation

The terms “route optimization” and “navigation” are often used interchangeably these days. Route optimization is really all about planning the best possible route, taking into account a number of variables. Navigation systems on the other hand, provide turn-by-turn directions to help drivers execute the route plan. In the case of dynamic systems, the route plan may change as the driver navigates the route and gathers real-time information about conditions ahead.

The new dynamics of routing

In the days of paper maps, routing information existed in a static format and the navigation process only became dynamic and interactive when you got lost and had to stop and ask for directions. Now electronic devices of all kinds, from computers to cell phones, are being deployed as interactive routing tools and even the maps themselves have become dynamic references that change along the way.

Today fleets and individual drivers can chose from integrated stand-alone routing and navigation systems; tools that integrate with wireless information systems and pull data from them into the routing process; third-party routing applications for laptop, handheld or dash-mounted computers; portable navigation devices and hosted systems for cell phones — and everything in between. Sharing so much routing information, via so many different devices in such a dynamic environment is not an easy task, however. One solution has been the development of so-called hybrid systems.

Hybrid routing systems deal with the problem of passing so much data back and forth by storing some routing information right on a navigation device such as an onboard computer and then serving other routing information from a host site to that device. In May, for example, QUALCOMM Inc. and Maptuit Corp. announced that they had entered into an agreement to make Maptuit's NaviGo available on QUALCOMM's OmniVision mobile computing platform.

NaviGo is a real-time hybrid system. It uses onboard and server-based technologies to deliver an in-cab navigation service designed to provide drivers with interactive maps. “Our hybrid solution allows drivers to send us feedback about route details or changes,” explains Luke Wachtel, business development manager for Maptuit. “This ability to ‘repaint’ the map in real-time as you go is the primary benefit of the hybrid approach.”

ALK Technologies is also exploring hybrid routing systems, according to Ed Siciliano, vp sales and marketing for ALK, but in a different way and for different reasons. “For the trucking world, we think it is a good idea to have the data live right on the device itself to save communication time and cost,” he says. “However, we believe that personal navigation devices such cell phones will eventually become the most popular platform. Those devices just don't have the computing and storage power necessary, so we are developing a Java application that will deliver routing information a little at a time to these devices. The system will essentially say, ‘I see where you are; I see where you are going,’ and then provide information a little bit at a time as the driver progresses along the route.”

“For some trucking companies, real-time information is essential, but for many others it is not,” says Kelly Frey, executive vp of Turnpike Global Technologies , provider of an electronic onboard data recorder (EOBR) and fleet management solution. “For example, for many fleets routing is primarily about planning not about finding your way through an unfamiliar area. Our EOBR can tell you what actually happened on a trip. Then you can use that real historical data to update and improve your route optimization plan,” he explains.

“We do fuel tax reporting, so we have to keep data for six years,” Frey continues. “That enables customers to look back at their history and ask questions such as, ‘how many times did we send two trucks to the same address on the same day?’

“We also collect GPS data points for our customers,” he adds. “Companies like Inrix can aggregate anonymous GPS data with that from thousands of other commercial vehicles; combine it with other traffic flow, event and weather data; and then deliver back road speed information. That enables companies to predict how long it will take to get across the border at a particular spot on a particular day, for instance.

“If a customer needs dynamic routing, we can feed all that information through a system like Cheetah , which will translate it into a recommended route and provide the information over a cell phone,” Frey notes. “We partner with Sprint/Nextel and QUALCOMM and we believe that combining cell phone-delivered navigation with an EOBR gives customers a route optimization solution at a lower cost that is also easier to deploy.”

Penske Logistics provides supply chain management and logistics services to major industrial and consumer companies throughout the world. According to Chuck Papa, vp information technology, access to so much historic and real-time data can be very valuable, but it puts new pressures on operations, as well. “More real-time data puts pressure on us to make sure that we have the right processes in place to manage all that,” says Papa. “Decision support structures, for example, must be fine-tuned for every customer. Sometimes that is a very involved process and sometimes it is not.

“When EOBRs are mandated, they will flood us with even more data, but they will not help us decide what to look at,” he adds. “When it comes to fleet management, we are at the stage where we have a wealth of information, but not a wealth of knowledge telling us what we should be acting on. That is the next stage in this evolutionary process.”