Technical information

Therefore, a battery charging isolation system was added including two additional heavy

density batteries. The additional batteries allowed more energy to be stored, while

driving to and from the work zone, and used during the short time use of the ACM

equipment. The system was tested repeatedly and found to be a sound solution.

The cone stowage assembly on the ACM was an area that required frequent inspection

and maintenance. Due to the increased interest in extended field testing, it was decided

to modify the design to increase the robustness of the assembly. There were two

components on the stowage assembly that required high maintenance, the v-track rollers

and the main radial-load flat track rollers. The flat track rollers were simply undersized

for the mechanical loading encountered and would be replaced with a larger one.

Therefore, a new gripper assembly had to be built to accommodate a larger roller. The v-

track rollers were splitting due to higher than expected torsional loading on the gripper

assembly. The addition of a third track providing support on the top of an existing flat

track roller greatly reduced the torsional loading on the v-track rollers. The two

improvements to the gripper assembly have improved the robustness of the stowage

assembly and greatly reduced the required maintenance.

Prior to continuing field testing, new cones were ordered, according to the latest Caltrans

specification, and installed on the machine. However prior to installation, the rubber feet

on the cones had to be removed, as required by the current ACM design. The new cones

were more pliable than the previous cone. The pliability affected several parts of the

cone placement and retrieval process; however, the detection and grabbing of the cone in

the secondary funnel area was problematic. The upper portion of the cone periodically

collapses, sliding under the secondary funnel, allowing the base of the cone to get trapped

behind the flag of a limit switch. This configuration prohibits the cone from being

detected or retrieved. Therefore, new flag geometry was developed that eliminated the

possibility of a cone being trapped and undetected. The new design was tested and

worked effectively.

Testing in Santa Cruz on highway 1 with Granite Rock Construction Company was an

excellent opportunity to observe the interaction of the ACM with congested traffic. The

distraction to drivers of having an interesting automated machine operating next to them

and the physical obstruction of the ACM drop box were two areas of concern to

investigate. It was found that neither of the concerns were problematic. Most drivers

never even noticed that the ACM was working autonomously. In addition, the protruding

drop box did not appear to be problematic.

Caltrans conducted further testing of the ACM in the Livermore and San Jose areas. On

the freeways the machine also performed great. The only real concern was the spacing

accuracy of the cones. This is a known problem and is limited by the physical speed at

which the ACM components can operate and the speed of the vehicle. With a little

practice the operators were able to speed up or slow down their driving to get a more

desired spacing.