Reducing Traffic Flow into Boston
Esteemed City Manager,
I urge you to consider the following ideas, in order to achieve the established objective of reducing traffic into our city by 20% over the next 5 years.
Theoretical Considerations:
In talking about creating a minimum viable product, it is important to first note its key characteristics. According to Eisenmann, Ries and Dillard in “Hypothesis-Driven Entrepreneurship”, each MVP must represent the smallest set of activities needed to disprove a hypothesis. They argue that the effort will be restricted in one or both of two ways: product functionality and/or operational capability. Thus, it either offers a limited set of features or relies on temporary technology (or both). It is advisable to carry out an MVP test with a constrained set of customers, when compared to the total population that a scaling enterprise would seek to service. Lastly, the authors caution against potential exposure to reputation risk. I would argue that this is important, given the extensive media coverage that often surrounds public sector projects, and the accountability that must be given to tax payer dollars.
In considering this exercise, it is also worthwhile to first list the limits to the lean startup methodology detailed by Eisenmann, Ries and Dillard. Firstly, low mistake acceptance is noteworthy in this scenario, since the mayor’s reputation (and that of his administration) might be seriously undermined via a notable failure. Secondly, it is advisable to avoid long development cycles that will prevent launching early and often. This immediately raises a red flag if we consider a multi-hundred million USD approach.
Test Layout and Structure:
Therefore, the approach selected must contain the following characteristics: short development cycle, discretion (to mitigate potential reputation risk), easy measurability, constraint to a smaller population within Boston, and limitation in features (and/or technology). The authors also note the value of parallel testing (applicable when relevant hypotheses are not serially dependent). Given the identified need for localized testing (say at the neighborhood level), I suggest a solution that can be tested simultaneously in several neighborhoods which suffer from varying levels of traffic intensity.
Given that the problem is framed as reducing traffic “into” the city, the solution should presumably target commuters, more than residents living within it (which could benefit from solutions such as electric scooters, bicycle lanes, etc). Autonomous vehicles have the vision of increasing vehicle usage efficiency via “Uber Pool style” commuting, which would maximize passenger seat utilization. I propose testing the impact of financial rewards in incentivizing drivers to maximize the commuting potential of their vehicles. The size of the cash reward could be associated with the commuting distance (correlated to the kilometers traveled) and the amount of passengers transported (bigger cars — resulting in greater savings — would be more rewarded). Volunteers can be identified separately, on a neighborhood basis, among suburban Boston communities, and offered different rates per kilometer, to gage the level of tradeoff while sacrificing their time. If done at the community level, the effort can be carried out discretely and the feedback loop will be fast, since results can be reevaluated on a weekly basis.
Limitations and Challenges:
The greatest obstacles are in the measurement and execution of the idea. I think that the sacrifice would need to be more in operational capability than product functionality, since the key data sources reside in verifying distance traveled with the number of passengers reported. A fully adequate tech solution would require significant investment — perhaps via an internally developed mobile app — so the MVP could be carried out using self-reporting. For example, different commuters could be required to send geotags of the start and end points of their commute, as well as a time stamped picture of the commuting vehicle’s license plate, to confirm its identity. For simplicity, these images could be uploaded via an API-enabled platform, or even sent via WhatsApp to a government designated number (the former is preferable).
Conclusion and Other Considerations:
Admittedly, this is a largely non-technical solution, and is quite based on civic engagement and cultural habits. Effectively, the incentivized driver is carrying out the role that an autonomous vehicle presumably would in the future. Depending on the estimated time frames for more sophisticated AV solutions to be rolled out, as well as the available budget funding, the investment of this MVP should be decided. Moreover, it should be again emphasized that the priority of the MVP is reducing traffic coming into the city (particularly during peak traffic times). It assumes that the majority of incoming traffic originates in suburbs; further research would be needed to validate this assertion. A similar approach could be carried out for returning commutes (from city to suburbs), yet that would be more difficult to implement, assuming that return times have greater variability than work entry hours. This is another point worth researching and validating.
