minor

slides/slides.tex

@@ -0,0 +1,153 @@
+\documentclass{beamer}
+\setbeamertemplate{navigation symbols}{}
+
+\usepackage{datetime2}
+\usepackage{booktabs}
+\usepackage{graphicx}
+
+\graphicspath{{../code/data/}{../code/model/}{../code/work/}{../images/}}
+\DeclareGraphicsExtensions{.pdf,.eps,.png}
+
+
+\begin{document}
+
+\begin{frame}[fragile]
+	\vspace{0.3\textheight}
+
+	\begin{center}
+		\large
+		On-demand public transport \\
+		\small
+		{\color{gray} is making us mobile }
+	\end{center}
+
+	\vspace{\stretch{100}}
+
+	\tiny
+	R.~Andreev \verb|<[email protected]>|, \today
+	\hfill
+	\tiny\color{lightgray}\hfill{\DTMnow}
+\end{frame}
+
+
+\begin{frame}
+	NYC TLC taxi data, May 2016:
+
+	{\ }
+
+	\resizebox{\textwidth}{!}{\input{../code/data/20210610-NYCTLC/e_explore/trip_table/green_tripdata_2016-05.tex}}
+
+	\vspace{2\baselineskip}
+
+	How many minibuses could meet the same demand?
+\end{frame}
+
+
+\begin{frame}[t]
+	From trip endpoints and travel time
+	we infer traffic conditions:
+
+	\begin{center}
+		\includegraphics[width=0.40\textwidth]{20210613-GraphWithLag/b_train/v1/lag/H=18/UTC-20210615-132827/train/velocity_i=9.png}
+	\end{center}
+
+	Use this for quickest routes.
+\end{frame}
+
+
+\begin{frame}[t]
+	Case study: 
+	\begin{itemize}
+	\item
+		100 requests during 18:00--18:17
+		within 1km of Times Square
+	\item
+		Allow -2min/+5min pickup and +10min dropoff windows
+	\item
+		Assume good traffic conditions
+	\end{itemize}
+
+	{\ }
+
+	Excess travel times (wrt quickest route) computed by optimizer:
+
+	\only<2>{%
+		\begin{center}
+			\small
+			10 single-passenger taxis
+
+			\includegraphics[width=0.8\textwidth]{20210616-OPT1/c_grid_study0/UTC-20210619-074952/c_grid_visualize/8/excess_travel_time_hist}
+		\end{center}
+	}%
+	\only<3>{%
+		\begin{center}
+			\small
+			10 minibuses of capacity 8
+			\\
+			\includegraphics[width=0.8\textwidth]{20210616-OPT1/c_grid_study0/UTC-20210619-074952/c_grid_visualize/9/excess_travel_time_hist}
+		\end{center}
+
+		Hence:
+		20 taxis $\approx$ 10 minibuses,
+		if allow detours.
+	}
+\end{frame}
+
+
+\begin{frame}[t]
+	Assume
+	\begin{itemize}
+	\item A certain proportion of minibus takers.
+	\item Excess travel times in dollars according to 2019 income census.
+	\item Prefer minibus if income/min $<$ expected excess travel time.
+	\end{itemize}
+
+	\only<2>{%
+		\begin{center}
+			\includegraphics[width=0.55\textwidth]{20210616-OPT1/c_grid_study1b/UTC-20210623-191629/e_evo_plots/evo__graph_h=6__graph_ttt_factor=3}
+		\end{center}	
+
+		\small
+
+		Assume bad traffic conditions (6pm traffic).
+		\\
+		Minibuses can meet only half of the demand.
+
+		\color{gray}
+		Faint: imposed fraction of bus takers.
+		Solid: net the unserviced requests.
+		\\
+		A = taxi price, B = minibus ride price
+	}%
+
+	\only<3>{%
+		\begin{center}
+			\includegraphics[width=0.55\textwidth]{20210616-OPT1/c_grid_study1/UTC-20210621-232339/e_evo_plots/evo}%
+		\end{center}
+
+		\small
+
+		Assume traffic improves (6pm $\leadsto$ 6am)
+		as more people take the minibus.
+		\\
+		Equilibrium at 50-90\% bus takers, depending on the price.
+	}
+\end{frame}
+
+\begin{frame}
+
+	\vspace{5\baselineskip}
+
+	The paradigm shift is feasible -- if it improves the traffic conditions.
+
+
+	\vspace{\stretch{100}}
+
+
+	\small
+	Details:
+	\url{http://bit.ly/optimum-2021}
+\end{frame}
+\end{document}
+
+