In conclusion, the AFOSR Aerothermodynamics portfolio hasworked closely with other agencies to envision and develop acomprehensive set of initiatives that have guided the strategy,performance and transition of essential science in support ofcurrent and future hypersonic technology development. Thisachievement would not be possible without the exceptionalcontributions of the outstanding researchers supported by theportfolio. Within this group, the students who are being preparedto lead the development of the next generation of hypersoniccapabilities represent the most important technology transitionfrom the portfolio.

Even with evolving global political and military challenges, theability to rapidly and economically cover large areas remains aninvariant goal for the warfighter. The objectives of the Air ForceOffice of Scientific Research Aerothermodynamics and Turbulenceportfolio are to provide the foundation for the achievement of this

Although individually the research highlights noted above areimpressive, when considered as a group it is the opinion of theauthor that they are absolutely game-changing. Once identified,knowledge of dominant energy transfer mechanisms can poten-tially be exploited to enable a revolutionary approach to thecontrol of macroscopic flow behavior. Specifically, flowfields couldbe designed to favor preferred energy transfer mechanisms thatresult in an application-optimized flow state.

Throughout the history of hypersonic technology developmentinvestments in essential supporting scientific efforts have beenclosely aligned with the technology system or application ofinterest at the time. In its 2000 report “Why and WhitherHypersonics Research in the US Air Force?” [32] the Air ForceScientific Advisory Board (SAB) described the roughly 15-yearboom-and-bust cycle hypersonic technology development hasendured in the United States and observed that as a result of theclose-coupling between technology and basic science funding thebase of scientific expertise in areas supporting hypersonic devel-opment has been slowly diminishing. The SAB observed thatexperts seeking alternative research areas during each bust cyclewere not being recovered or replaced with each new boom period.In 2003 program managers from AFOSR, NASA Langley andSandia National Laboratories recognized the fact that, while eachorganization had a unique mission and technology objective, similarfoundational science capabilities and investments were required. Asa result, efforts began to coordinate the research investments ofeach agency in the area of laminar–turbulent transition with thegoal of promoting collaboration and maximizing the limitedresources available. By 2005 the effort has seen sufficient prelimin-ary success that, when DoD senior leadership suggested that anational initiative coordinating hypersonic research would be well-received, it was adopted as the basic model for the NationalHypersonic Foundational Research Plan. Although there have beenseveral prior efforts to coordinate hypersonics research on anational scale, the intent of the National Hypersonic FoundationalResearch Plan was to employ a slightly different approach byfocusing on scientific, not technology, challenges.

However, over the last five years there has been asubstantial increase in the number of technology transition eventsderived from the STAR effort, as the team's expertise and methodshave continued to consistently make valuable contributions to theresolution of key scientific challenges

several workshops to inform the technologydevelopment community with regard to the maturing basicresearch tools that were potentially applicable to the Falconprogram and later consulted with various groups on technicalissues related to laminar–turbulent transition. As the Falconprogram progressed the STAR team provided critical insighttowards the verification of the vehicle configuration and trajectoryand helped identify key events and phenomena that contributedto in-flight incidents. In addition to supporting the Falcon HTV-2program the STAR team also contributed to the resolution of keychallenges for the X-51 program and the transition of methodsdeveloped by the team has facilitated unprecedented new cap-abilities within the Test & Evaluation community.

In summary, the HIFiRE program has played a considerable rolein extending the legacy of HyShot and HyCAUSE and shaping thecurrent philosophy towards economically-efficient flight research.The success rate of the HIFiRE flights, as well as those from the DLRSHEFEX [21] program which shares a similar philosophical foun-dation with HIFiRE, have led to the appearance of smaller-scalerisk mitigation flight research components in the current genera-tion of major hypersonic technology demonstration programs suchas the proposed DARPA IH program. Additionally, flight researchprograms led by academia, such as the University of Virginia/Virginia Tech High-V program [22] are emerging.

The 80% success rate for the program so far ismuch higher than that typically achieved for a flight researchprogram and is a credit to the Program Manager, Mr. DouglasDolvin from AFRL, and the rest of the HIFiRE team

In this role, AFOSR emphasizes the identificationand support of innovative, unique research that has the potentialto provide game-changing capabilities for the future Americanwarfighter. The AFOSR Aerothermodynamics & Turbulence portfo-lio has responsibility for fluid dynamics research associated withhigh-speed and high-energy flows of interest to the Air Force.Current emphasis areas within the portfolio include the funda-mental physics of turbulence and boundary layers, shock-dominated flows – especially shock/boundary layer and shock–shock interactions, and flows in thermochemical nonequilibrium.The dual objectives of the Aerothermodynamics & Turbulenceportfolio are to identify and foster the development of innovativescience with the potential to lead to transformative new capabil-ities while simultaneously championing the technology transfer ofresearch breakthroughs to application within technology matura-tion programs.

dominant challenge for the group of gliding systems compos-ing the buff-colored region in Fig. 2. For these systems rate-dependent thermochemical processes determine the release ofenergy from excited internal states into thermal energy and theaccurate prediction of aerothermodynamic phenomena is pacedby the accuracy of relevant reaction rates required for simulations.

Thus,the ability to efficiently and accurately predict the aerothermody-namic base state becomes a key driver for the advancement ofsystem design and analysis capabilities

a variety of scientific disciplines including fluiddynamics, thermophysics, high-temperature materials, chemistry,and computational science.

Superior speed and range have long been recognized by militaryscholars as game-changing advantages and the benefits of hyper-sonic capabilities have inspired us in this manner for almost half acentury.

The last decade has witnessed incredible advancements in thetechnology and foundational scientific knowledge base essentialfor the development of efficient future hypersonic capabilities. Thebasic research community has played a vital role in the develop-ment and transition to application of innovative methods andfoundational insight that have guided progress, and in this arearesearchers sponsored by the Air Force Office of Scientific Research(AFOSR) have been at the forefront

A fully electrified fleet is not advantageous to the operator right now, but public EV charging stations are becoming more widelyavailable. EVs are becoming less expensive to own and operate, and the future of fossil fuels is not clear. The cost to run this EV fleet isstill quite low on a per-mileage basis—less than driving a personal vehicle 10,000 miles per year (AAA, 2015) for the low- and mid-range cost estimates. It is good to know there are alternatives to fossil fuels that can be profitable for such a fleet with the uncertainfuture of our climate and fossil fuel prices.

Vehicle costs were estimated based on small production EVs, such as the 2016 Chevrolet Spark EV and 2016 smart fortwo electricdrive coupe, with all-electric ranges (AERs) of 82 and 68 miles, respectively (U.S. DOE, 2017). These ranges are not far from the 60-mile assumption for short-range SAEVs used here. These two models have MSRPs of $25,120 and $25,000 respectively (U.S. DOE,2017). As for long-range EVs, the 2019 Chevrolet Bolt has a 238-mile range and costs $36,620 (Chevrolet, 2019). Tesla begandelivering units of their new Model 3 in 2018, selling at $44,000 before any subsidies and with a range of 303 miles (Tesla Motors,2016, 2019). These prices do not include government rebates, which are due to be phased out in the near future (IRS, 2016), soshould not be depended upon for this study. Vehicle autonomy is reported by ENO (2013) to have an estimated marginal cost of$25,000 to $50,000 but this cost could come down to $10,000 after at least 10 years. For this analysis it is assumed that autonomywill have a marginal cost of $5000 to $25,000, and that regular range SAEV, without autonomy will cost $25,000 and a long rangeSAEV will be $35,000. With the autonomy package this gives prices of $30,000 to $50,000 for short range SAEVs and $40,000 to$60,000 for long range SAEVs. The cost of HEVs is estimated as $20,000 without autonomy.Similar to Chen et al. (2016), SAEVs are anticipated to last 215,000 miles, similar to the average lifespan of a NYC taxicab (NewYork City Taxi & Limousine Commission, 2014). Life cycles of such rigorously used EV fleets have not been studied and may havebetter or worse lifespans. A battery's usable life is estimated at roughly 100,000 miles based on standard practice by OEMs towarranty their batteries for this distance plus various reports such as Saxton (2013). Then a battery will need to be replaced at leastonce during a vehicle's lifetime, but it would not be a good investment to replace the battery a second time since the vehicle will bevery close to (if not in excess of) the end of its service-life. Replacement batteries are expected to cost between $100 and $190 perkWh per estimates from GM and Tesla (Voelcker, 2016), substantially lower than recent estimates of $268/kWh in 2015 and $1000/kWh in 2008 (IEA, 2016). It's assumed that a trained technician could replace a battery in about an hour billing $50 an hour.Vehicle operation and maintenance costs are assumed similar to those for conventional, privately-owned gasoline vehicles, whichAAA (2015) estimates to be 5.4 to 6.6 cents per mile for various vehicle types. Cleaning is difficult to estimate, in part becausecleanliness standards are subject to operator judgement, and operators may impose stiff penalties for travelers soiling or damagingtheir vehicles (as is quite common with today’s TNC services [see for example https://www.ridesharingdriver.com/uber-fees-cancellation-booking-cleaning-fees/]). Bösch et al. (2017) estimate that a 4-passenger SAV will need to be cleaned every 40 revenue-trips at a cost of $15.26 per cleaning. Since cleaning costs could be very low (or even profitable, thanks to damage fees), those costsare added here only to the high-cost estimates (at an overall rate of 2.6 cents per occupied-mile

ion’s outcomes.As one might expect, the results indicate that the HEV fleet serves travelers the best, rejecting only 1.62% of trips and meetingtrips with an average response time of 4.45 min. Also, not surprisingly, the low-range, slow-charge SAEV fleet served travelers theworst, rejecting 55% of trips due to poor response time and another 5.4% on the basis of trip length, leading to a vehicle replacementrate of only 3.75. T

At this point, it might be intuitive to drop off Traveler A first. However, travel times between Origin C and Destination A are not inthe MATSim trip file, and so must be estimated through teleportation. The trip between Origin C and Destination C is in the trip file, sothat travel time is well estimated through traffic assignment. Therefore, in order to preserve the highest degree of realism, the lasttraveler picked up must be the first to be dropped off. This may appear unfair or inefficient, but the algorithm ensures no travelerexperiences a delay greater than 20% to his/her in-vehicle travel time (IVTT), so travelers A and B will not be too inconvenienced.Travelers will always share rides if doing so minimizes response time. Finally, no more than four travelers may share a vehicle

Many of the trips produced in the MATSim trip file are not reasonably serviceable by the SAV fleet due to their spatial dis-tribution, trip length, or other factors that lead to traveler wait times of tens of minutes or even hours. In former uses of this model,Bösch et al. (2016) rejected requests when they were in the system for more than 10 min. Loeb et al. (2018) would reject any requestin excess of 75 km (46.6 miles). Unfortunately, neither of these models has any kind of stochastic behavior and does not acknowledgethat trips longer than 75 km may have short wait times or that many travelers may be willing to wait longer than 10 min. This isimportant for cost calculations, since a flexible demand model is necessary to understand how levels of service affect use levels andresulting effects on aggregated costs

This financial study is carried out using a simulation of a SAEV fleet across the Austin, Texas, 6-county region. The region’s travelpatterns are not unusual among Americans, and the area has a very similar density and size to many other U.S. regions, includingOrlando, Florida, Columbus, Ohio, and Milwaukee, Wisconsin. One can expect this paper’s simulation results applicable for a varietyof other regions, including the relative trade-offs between various fleet parameters, for regions of different size and/or density.

These surprising findings are likely thanks to the highly simplified and unrealistic model they employed, which resulted in highlyoptimistic results.

Theyalso assumed the fleet operator will be responsible for costs associated with owning and maintaining chargers, in addition to thevehicles.

This study simulates various cost scenarios using the data found in Loeb et al. (2018) to help a fleet operator determine if an SAEVfleet is a wise and feasible option, what charge speeds and range are the most reasonable and financially advantageous, and howthese results compare to simulations of an all-gasoline fleet.

Any self-driving fleet will incur high fixed costs, at least in early stages of the technology’s release, so scenarios under which sucha fleet is cost-effective, compared to a gasoline-powered fleet, should be explored before making this large capital investment, grantedsuch scenarios even exist.

may

A system of shared autonomous electric vehicles (SAEVs) can carry a relatively high fixed cost due to smaller scale productionand the cost of large batteries, which provide greater range before charging is required, and additional charging infrastructure, butmay reduce overall costs via lower energy and maintenance needs.