The Deployment Tracking survey began in 1997, with it’s modern form emerging in 1999. Survey years include 1997, 1999, 2000, 2002, 2004, 2005, 2006, 2007, and 2010. The next survey is planned for 2013

The gaps are years in which the survey was not administered.

The gaps are years in which the survey was not administered.

The Volpe Center conducted a thorough quantitative examination of the deployment tracking data from 1999 to 2007. The goal was to begin to understand t his deployment information in light of some of the policy levers that could be used to influence adoption or deployment.

The markets examined included ETC, HDC, TSP, TMS, EVP, and VDC. Mature markets are those that are saturated in terms of deployment. Growth in deployment and adoption have slowed significantly or plateaued entirely.

The average price of a cellphone was about $4,000 in 1984 – and only a few people could afford one.  Sales of home phones fell as cellphones got less expensive, averaging about $200 in 2000.  Sales of standard cellphones began to fall as smartphones added features beyond calls and text messaging.

These data show and compare the effect of an increase in budget or implementation of a regional architecture on the adoption of either EVP or VDC ITS technologies.

 

The bottom section of each of the charts show the baseline predicted probability of adoption for the median agency in the data sample (which is from the Deployment Tracking survey). In the case of EVP the baseline is 15% and for VDC it is 90%.  The probability of adoption for the median agency is considerably different in these markets.

The top section of the charts show the increase in adoption from increasing funding from the median agency level to the 99th percentile (in term of funding).  For architecture it is the effect of going from no architecture to having an architecture. 

A key observation here is that increasing either funding or implementing a regional architecture has similar effects on the magnitude of the subsequent increase in the probability of adoption for each technology.

For the purposes of this modeling exercise, the increase in budget for the median agency in EVP-FR is similar to giving Miami’s budget to Syracuse. Likewise, for VDC it is like taking the budget for Dallas and giving it to Birmingham, AL. (Note these are county level budget data, but they do provide and indication of the magnitude of funding change required to shift the adoption probability by the percentage noted above – the way to interpret these data is that the increase in probability is additive to the baseline). 

If the cost of implementing a regional architecture for the median agency is considerably less than the cost of funding  noted above, then a regional architecture will provide close to same increase in adoption with a much lower outlay.  

Another observation is the magnitude of change in the adoption probability given the baseline predicted probability of the median agency.  The higher this probability, the less effect an intervention, be it funding or an architecture, will have.

If the cost of additional funds to increase adoption is considerably larger than the cost of implementing or upgrading a regional archaitecture requirement, then 

 

  

This stylized diffusion curve highlights how to combine the findings from the Volpe Center report. In the early stages, technology adoption will be slow. This can be helped along by interventions such as PCB and KTT which spread the word about a given technology, potentially reaching those agencies which are imitators. After the word has gotten out, direct interventions can be effective and tip the scales in favor of adoption. Finally, as technologies reach maturity, there is little impact to be had via direct intervention; the majority of potential adopters have already adopted. Here, the imitation effects that initially slowed deployment are dominant.