16-Civ-A6 Highway Design, Construction, and Maintenance · May 2013
Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)
Paper format. National Examination, 98‑Civ‑A6 Transportation Planning & Engineering (May 2013). Closed book, one two‑sided aid sheet, 3 hours. Seven questions; any five constitute a complete examination and each is of equal value (20 marks). All seven are solved below as a study resource.
Reference texts (subject).
Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.
This is a conceptual question spanning the land‑use/transport feedback cycle, a known theoretical weakness of the multinomial logit model, and the classic policy dichotomy between adding capacity and managing demand. Each part is answered as flowing prose.
Land use and transportation are bound together in a continuous two‑way feedback loop, often drawn as the “land‑use/transport cycle.” A concrete example is the construction of a new rapid‑transit station (or a freeway interchange) at the edge of a city. The improved accessibility that the facility creates raises the development potential of the surrounding parcels, so over the following years those parcels attract higher‑density residential, retail and office development. That new development is itself a generator and attractor of trips.
The effect on travel demand is therefore both immediate and long‑run. In the short run the accessibility improvement lowers the generalized cost of reaching the area, which increases the number of trips destined there and can shift trips from other destinations and modes. In the long run the intensified land use permanently raises trip generation and attraction rates in that zone, feeding still more travel demand back onto the network — the mechanism by which new highway capacity can eventually “fill up” through induced development. Because activity location responds to accessibility and accessibility responds to activity location, transportation planning cannot treat land‑use patterns as fixed inputs; the two must be forecast jointly.
The multinomial logit model is derived by assuming that the random (unobserved) components of the route utilities are independent and identically distributed Gumbel variates. Independence of these error terms produces the property known as the Independence of Irrelevant Alternatives (IIA): the ratio of the choice probabilities of any two routes depends only on those two routes’ own utilities and is unaffected by the presence or attributes of any other route.
When two or more routes physically overlap (share common links), that assumption is violated. Overlapping routes are not perceived by travellers as fully distinct alternatives — their travel times are correlated because they share the same congested links, so their unobserved utility components are correlated rather than independent. The classic illustration is the “red‑bus/blue‑bus” paradox transferred to routes: if one physical path is represented as two nearly identical overlapping routes, the MNL model treats them as two independent choices and jointly captures far more traffic than the single physical corridor should, unrealistically penalizing a genuinely distinct alternative.
The limitation is overcome by relaxing the IIA assumption or by correcting for the overlap. Practical remedies include: (i) a path‑size logit or C‑logit model, which adds a “commonality factor” (a correction term measuring how much a route shares links with others) to each route’s utility so that overlapping routes are down‑weighted; (ii) a nested logit model, which groups correlated (overlapping) routes into a common nest so that the correlation is captured by the nest structure; and (iii) a probit or mixed‑logit assignment, which allows a full covariance matrix among route utilities and so models the shared‑link correlation directly. Any of these restores realistic splits between overlapping and independent routes.
A supply‑side solution attacks a transportation problem by increasing the capacity or performance of the transportation system itself — it changes what the network can supply. A demand‑side solution instead manages, reduces or re‑times the travel demand placed on the system, so that the existing supply is used more efficiently.
An example of a supply‑side solution is widening a congested arterial by adding a lane, building a new transit line, or improving signal timing and intersection geometry to raise throughput. An example of a demand‑side solution is a travel‑demand‑management (TDM) measure such as congestion (road) pricing, parking pricing, staggered work hours, or promotion of carpooling and telecommuting, all of which lower or spread the peak demand rather than enlarging the road. In practice the two are complementary: because of the land‑use feedback described in part (a), supply‑side expansion alone can induce new demand, so durable congestion relief usually pairs capacity with demand management.
