On July 20th, Ross Sheckler of iCone made a presentation to the Autonomous Vehicles Symposium in San Francisco. The title of his presentation was “What Do Automated and Connected Vehicles Need to Know About Work Zones?” His message was very important. It was well-received by those in attendance, but the group that needs to hear this is many times larger than the 100 or so people in the room that day. So we will try to make his main points in today’s post.
Remember, most of the attendees were not work zone people, though a few of us were there that day. Most work for automotive manufacturers or component manufacturers. They produce navigation systems – some in use today and some that will guide autonomous vehicles in the future. Those cars will drive through our work zones, yet the folks who produce them know very little about temporary traffic control. So Ross began by pointing out that the map changes 1,000 times per day due to work zones. 1,000 times per day workers change the law, and 10,000 times per day warnings are posted. His point being, of course, that we must find a way to inform these systems.
Mr. Sheckler also explained that most closures are never reported. And of those that are reported, most don’t occur on the dates and times they are scheduled. He went on to say that the most dangerous closures are probably those unreported ones. He used the example of a short term utility closure on a rural road with bad line of sight. The people doing that type of work often do not worry much about traffic control. They might place a 10 foot taper of cones and a ROAD WORK AHEAD sign, but even that is somewhat rare. Automotive systems must be able to recognize these work areas and react appropriately.
And when traffic control is reported, it only shows up in navigation apps as “roadwork”. It does not say it is a lane shift, or multiple lane closure sure to cause queuing. It does not say the entire geometry has changed by moving traffic over into the oncoming lanes separated by concrete barrier. And it does not tell you if the work is causing traffic to slow or stop. A shoulder closure is reported the same way as a full roadway closure with detour. Yet one does not affect traffic at all while the other may affect travelers’ choice of routes.
His point is that by reporting these changes as they occur it gives drivers the opportunity to avoid the area altogether. But the information must be posted as the changes occur and it must be accurate. If it is, drivers will learn to depend on it and change their routes. But if they get erroneous or inaccurate information, they will continue to drive along their intended path.
Ross finished by listing the details that are important to navigation apps, and this applies to current apps as well as future autonomous driving systems.
- Work zone status: scheduled versus equipment on sight and ready to work versus workers present.
- Map changes including lane shifts, capacity reductions of any kind, or roads closed.
- Queue details including slow or stopped traffic, delay times, early or late merge systems, and location of merge point.
- Presence of active flagging operations including location.
- Presence and location of attenuator trucks, especially when the attenuator is in the down or active position.
These are all details a system will require to make informed routing recommendations. And if the work does cause significant impacts, we prefer they avoid the area altogether. It is safer and more efficient for everyone involved: travelers, contractors, and for the owner/agency.
Our industry can supply this information today. So please encourage system designers to engineer with that in mind. We can all avoid a future full of expensive, time consuming, and even dangerous problems by getting the word out now.