In the map of Figure 1, the homes along Pine Tree Drive are much more expensive and, thus, potentially of heavier significance for urban-scenic routing, than those along the shortest route. A slightly longer drive or walk, with property values taken into account for urban-scenic routing, would be along Pine Tree Drive, as shown in Figure 2.
If lesser, but not insignificant, weight is assigned to the urban-scenic interest, then the route would be slightly shorter, yet still longer than the shortest route, as shown in Figure 3 .
The relative importance of time, cost of travel, and urban-scenic interest can be determined by the user utilizing the Ullrich et al.[5] method of a weights-selection triangle: a touchable triangle that allows the user to assign importance weights to three interrelated decision optimization objectives using a single gesture, as shown in Figure 4 et seq.
Applying the Ullrich et al. method[5] to the weighting selection problem modeled here, three objectives (A=time, B=cost of travel, and C=urban-scenic interest) are presented in a triangular fashion on a touch screen. Figure 4 shows the underlying principle of the establishment of a single weight wA for Objective A; Figure 5 combines three objectives into a single triangle, allowing for the establishment of a tri-variable weight function (wA, wB, wC). By applying a finger gesture, the user moves an indicator freely inside the triangle (see Figure 6). The position of the indicator establishes a tri-variable weight function, which in further steps is then used as input for a co-optimization algorithm. When the user is satisfied with the established weights, she indicates this, e.g., by pressing a touch screen button labeled “Go”.
The mere consideration of property values might include properties of the kind that the user does not consider worthy of observing on her trip, e.g., commercial properties. The user may narrow down the property values to be considered in the weighting algorithm to be restricted to certain categories of homes. For example, in the selection criteria of Figure 7, the user can choose between various property types and select, e.g., only the type of single-family homes.
The routing may be changed to remove, for example, commercial properties and multi-family residences from contributing to the urban-scenic routing criterion (see Figure 8).
The user may include arbitrarily complex criteria for inclusion or exclusion of types of properties in evaluating the urban-scenic interest. For example, the user may choose to select only single-family homes with a lot size of at least 10,000 sq. feet, as in Figure 9.
With certain weights attached to the various criteria, the route may be like the one shown in Figure 10.
The routing can be presented to the user via oral instructions, in a graphic form, or in textual form, as shown in Figure 11.
The weighting desired by the user might be based on the total dollar value of the home or on another related metric, which might better capture the user’s needs. For example: the home value per square foot, as in Figure 12.
Rather than the entire home value, the value for weighting might be just the value of the structure (what in real estate is called the “improvements”) or the value of the land without the structure (the “unimproved land” value). An example with just the values of the unimproved land is shown in Figure 13.
The source of the valuation of each home can be, for example, the assessed value of the home per county records, recent sale price, or current asking price from the MLS records. In the case of a county assessed value, one would typically choose for the purpose of the herein presented weighting an objective value rather than tax valuation, since the latter may be dependent on the property owner’s status rather than only on the objective property quality. For example, in Florida, counties publish multiple “values” for the same home, including “the taxable value,” i.e., the value against which the property tax is assessed and which takes into account the freezing of homestead property valuation and various discounts to which the current property owner may be entitled. A more objective county-published value in Florida in what the counties call the “Just Value.” While it may or may not be a true reflection of the current value of the property, it is objective in the sense that the county applies the same methodology to estimate the “just values” of various properties; thus, it can be useful for the weighting presented herein (see Figure 14).
The example shown in Figure 15 shows the various official “valuations” available from Florida counties. Among these valuations, the most meaningful for urban-scenic routing purposes is the “Just Value,” while the Land-value and Building-value are also meaningful. The other valuations are affected by the demographics of the property owner and, thus, are not meaningful for urban-scenic routing purposes.
Other objective metrics can be computed utilizing the published data. For example, the value per square foot can be computed from the published home value and the published home size, as in Figure 16.
Figure 17 presents is a different example of a source of home values: the current asking price in the real-estate multiple-listing services (MLS).
When the source property value is per-house, it can be translated into value weight per street segment using an appropriate statistical aggregation of data. The example in Figure 18 shows the computation of the Maximum and the Average home value along the 4200 segment of Sheridan Avenue.
Many other reasonable statistical aggregation functions for the purpose of urban-scenic routing include:
- Median value
- Average value after exclusion of low outliers
- Median of the highest 20% of values
- The number of homes valued at over $1M
- The number of homes valued at $1 to $2M plus double the number of homes valued at over $2M
While the aforementioned examples considered sourcing property valuation per house and then their aggregation per street segment using various statistical methods, another application of the herein present method may use already pre-aggregated property values as may be available, for example, in the United States from the American Community Survey (ACS) or the United States Census (Census). However, the mentioned data source examples may have a sparser spatial granularity than a street segment, in which case the urban-scenic routing method would be slightly less precise: for example, the blocks 4200 Sheridan Ave and 4200 Pine Tree Dr are within the same home valuation statistical area in ACS, and they are in the same block group in Census. Furthermore, one street segment may lie on the boundary of two statistical areas, in which case the urban-scenic valuation of the street segment should combine the even side of the street segment and the odd side of the street segment.
After the home values have been aggregated per street segment using, for example, any of the aforementioned per-house or sparser data sources, the aggregated values need to be normalized over the entire relevant map portion. For example, the aggregated values can be normalized into the range of 0 to 1. Thereafter the total normalized value of each street segment can be computed by considering said normalized values in conjunction with other criteria, such as the street segment’s expected travel time. Relative weights are assigned to the various criteria, using, for example, the aforementioned touch triangle method.
Once a route is computed, it can be presented to the user for approval.
If the routing is presented to the user in a graphic form, it may be further enhanced in various visual forms to inform the user and let the user visually confirm that the choice of the route shows what the user intended or have the user adjust the relative weights and criteria. For example, overhead imagery of the houses that the user would pass by (on the currently offered route) can be displayed, as in Figure 19.
Another way is to display photographs of the facades of the houses that the user will encounter, as in Figure 20.
Another example is to present to the user oblique (bird’s eye view) images as shown in Figure 21.