Recent advance­ments in infor­ma­tion and vehic­u­lar tech­nolo­gies dri­ve the wave of inno­va­tions in mobil­i­ty ser­vices and sys­tems. The fig­ure on the right presents the time­line of key tech­nol­o­gy devel­op­ments and the emer­gence of new mobil­i­ty ser­vices, and shows how the lat­ter fol­lows the for­mer. Specif­i­cal­ly, the num­ber of smart mobile devices in the US has been ris­ing steadi­ly. These devices retrieve users’ geolo­ca­tions, enable ubiq­ui­tous com­mu­ni­ca­tions, and allow instant peer-to-peer inter­ac­tion, giv­ing rise to var­i­ous on-demand mobil­i­ty ser­vices for goods and peo­ple, which bring togeth­er sup­pli­ers of resources (e.g., car and park­ing space) and ser­vices (e.g., ride and park­ing) with very low trans­ac­tion costs. Con­nect­ed and auto­mat­ed vehi­cle tech­nol­o­gy will fur­ther rev­o­lu­tion­ize urban and rur­al mobil­i­ty and pro­mote the shift from car own­er­ship to sharing/subscription. Auto­mat­ed shared-use mobil­i­ty ser­vices may even­tu­al­ly emerge where com­pa­nies own a fleet of dif­fer­ent types of auto­mat­ed vehi­cles and offer on-demand ride hail­ing ser­vices. Oth­er types of mobil­i­ty ser­vices will be like­ly cat­alyzed by new busi­ness mod­els and ven­ture cap­i­tal investments.

These mobil­i­ty ser­vices are expect­ed to play an increas­ing­ly impor­tant role in meet­ing mobil­i­ty needs. It is crit­i­cal to under­stand the impacts and impli­ca­tions of these ser­vices and pro­vide guid­ance on their devel­op­ment and deploy­ment. LIMOS con­ducts quan­ti­ta­tive analy­ses and devel­ops nov­el mod­els and tools to under­stand, plan, design, and reg­u­late emerg­ing mobil­i­ty ser­vices to enable them to be inte­gral com­po­nents of trans­porta­tion sys­tems and improve sys­tem con­nec­tions and inte­gra­tion. Below we high­light select­ed LIMOS research projects towards con­nect­ed, auto­mat­ed, shared and elec­tri­fied (CASE) mobil­i­ty.

Advanced Parking Management for Traffic Congestion Mitigation

Park­ing is a grow­ing prob­lem in many dense urban dis­tricts. To many, find­ing a park­ing space in those areas is an unpleas­ant expe­ri­ence of uncer­tain­ty and frus­tra­tion. Cruis­ing for park­ing makes already-con­gest­ed urban streets even worse and yields sig­nif­i­cant waste in time and fuel. The pro­lif­er­a­tion of advanced smart­phones pro­vides tremen­dous oppor­tu­ni­ty for effi­cient park­ing man­age­ment.  Smart­phone-based park­ing man­age­ment appli­ca­tions have emerged. These appli­ca­tions help dri­vers find park­ing spaces by allow­ing them to use smart­phones to view real-time avail­abil­i­ty and prices of park­ing spaces and guid­ing them to open park­ing spaces, reserved or otherwise.

Fund­ed by Nation­al Sci­ence Foun­da­tion, we con­duct a com­pre­hen­sive study on smart­phone-based park­ing man­age­ment. The research gen­er­ates a set of ana­lyt­i­cal tools that not only explain the under­ly­ing work­ing mech­a­nism of advanced park­ing man­age­ment ser­vices with much enhanced rep­re­sen­ta­tion of traf­fic and behav­ioral real­ism, but also gauge their poten­tials for reduc­ing traf­fic con­ges­tion. The ana­lyt­i­cal tools lay out a blue­print for opti­miz­ing their imple­men­ta­tion and designs. The the­o­ret­i­cal efforts are com­ple­ment­ed by an agent-based sim­u­la­tion, which tests the valid­i­ty and applic­a­bil­i­ty of the the­o­ries, and unveils com­plex out­comes of park­ing com­pe­ti­tion under real­is­tic park­ing search behav­iors. The research advances the knowl­edge and analy­sis of park­ing man­age­ment and enrich­es the lit­er­a­ture of mod­el­ing morn­ing com­mute, vehi­cle rout­ing and game the­o­ry. Final­ly, the work sheds light on design­ing oth­er appli­ca­tions enabled by advanced smart­phone tech­nolo­gies for traf­fic man­age­ment and con­trol. [Learn more…]

Participatory Traffic Control

In many years to come, the traf­fic stream will be mixed with human-dri­ven vehi­cles with­out con­nec­tiv­i­ty (reg­u­lar vehi­cles), human-dri­ven vehi­cles with con­nec­tiv­i­ty (con­nect­ed vehi­cles), and vehi­cles with var­i­ous lev­els of automa­tion. We attempt to lever­age con­nect­ed vehi­cles in the traf­fic stream to bet­ter man­age and oper­ate our road net­works. More specif­i­cal­ly, we pro­pose par­tic­i­pa­to­ry traf­fic con­trol, where a com­mu­ni­ty of con­nect­ed vehi­cles will be incen­tivized to opt in for traf­fic con­trol and man­age­ment. We envi­sion that an online and mobile plat­form will tar­get spe­cif­ic par­tic­i­pat­ing com­muters and incen­tivize them to a) behave as “trav­el demand dis­trib­u­tors” to bet­ter dis­trib­ute the com­mut­ing demand across time peri­ods and trans­porta­tion facil­i­ties; and b) func­tion as “traf­fic stream reg­u­la­tors” to reg­u­late traf­fic stream to pre­vent or delay the acti­va­tion of recur­rent bot­tle­necks. Our work­ing hypoth­e­sis is that, by incen­tiviz­ing the behav­ioral changes of a small num­ber of tar­get­ed par­tic­i­pants, the plat­form can influ­ence a larg­er num­ber of untar­get­ed com­muters’ trav­el deci­sions to improve the over­all sys­tem per­for­mance. The over­ar­ch­ing goal of par­tic­i­pa­to­ry traf­fic con­trol is to migrate many func­tions of tra­di­tion­al, imper­son­al, phys­i­cal con­trollers in the trans­porta­tion sys­tem (e.g., mes­sage signs, speed lim­its) to the tar­get­ed behav­ior inter­ven­tion in the cyber space.

Fund­ed by ARPA‑E of the US Depart­ment of Ener­gy, we have worked with researchers from Uni­ver­si­ty of Mary­land to apply the above idea to reduce ener­gy con­sump­tion in per­son­al trans­porta­tion. The project yield­ed Incen­Trip (, an app that aims to reduce ener­gy use of vehic­u­lar trav­els by incen­tiviz­ing indi­vid­ual trav­el­ers to adjust trav­el choic­es and dri­ving behav­iors. The pri­ma­ry con­tri­bu­tion of LIMOS to Incen­Trip was to devel­op its con­trol opti­miz­er that opti­mizes per­son­al­ized award to encour­age behav­ioral changes.


Adapting Land Use and Infrastructure for Automated Driving

Auto­mat­ed vehi­cles (AVs) will like­ly yield a trans­for­ma­tion of urban mobil­i­ty sys­tems, which will fur­ther impact the form of a city and its land use, and sub­se­quent­ly, the trav­el pat­tern of the city. There­fore, it is crit­i­cal to inves­ti­gate these impacts and impli­ca­tions, which will help plan­ning and devel­op­ment agen­cies to form their poli­cies, reg­u­la­tions and plans for urban land use to sup­port auto­mat­ed mobil­i­ty. It is also crit­i­cal for plan­ning agen­cies to mod­i­fy trans­porta­tion infra­struc­ture to adapt to, and more impor­tant­ly, pro­mote the deploy­ment of AVs. Before man­u­al dri­ving can be crim­i­nal­ized one day as some have pre­dict­ed, the traf­fic stream on road net­works will still be het­ero­ge­neous, with both con­ven­tion­al vehi­cles and AVs. We envi­sion that plan­ning agen­cies can ini­tial­ly iden­ti­fy crit­i­cal loca­tions to imple­ment var­i­ous AV mobil­i­ty appli­ca­tions. For exam­ple, a “bot­tle­neck man­ag­er” can be imple­ment­ed at a recur­rent free­way bot­tle­neck. When approach­ing, AVs send requests via vehi­cle-to-infra­struc­ture wire­less com­mu­ni­ca­tions to the “bot­tle­neck man­ag­er”, which will pri­or­i­tize the requests and opti­mize their tra­jec­to­ries to ensure time­ly pas­sage while pre­vent­ing the bot­tle­neck from being acti­vat­ed. To lever­age the grow­ing adop­tion of AVs, plan­ning agen­cies may lat­er change the lane con­fig­u­ra­tion of cer­tain high­way seg­ments to ded­i­cate some traf­fic lanes to AVs only to facil­i­tate the for­mu­la­tion of vehi­cle pla­toons for high­er through­put. Some lanes may be ded­i­cat­ed for pas­sen­gers’ pick-up/­drop-off. Park­ing spaces in busi­ness dis­tricts will be relo­cat­ed to the periph­ery. The relo­ca­tion of on-street park­ing may pro­vide more lanes and increase road­way capac­i­ty in busi­ness dis­tricts. More­over, the AV-only lanes can be fur­ther expand­ed to AV areas, and sub­se­quent­ly imple­ment­ed are inno­v­a­tive con­trol strate­gies that aim to achieve sys­tem opti­mum per­for­mance in those areas. The ded­i­cat­ed AV areas will fur­ther expand grad­u­al­ly as the lev­el of mar­ket pen­e­tra­tion of AVs increas­es, and even­tu­al­ly sup­port ful­ly con­nect­ed and auto­mat­ed mobil­i­ty in the whole system.

Var­i­ous State Depart­ments of Trans­porta­tion and Met­ro­pol­i­tan Plan­ning Agen­cies around the nation have rec­og­nized the need for land use and infra­struc­ture adap­ta­tion plan­ning for AVs. There is a lack of sys­tem­at­ic method­ol­o­gy to sup­port such pol­i­cy mak­ing and plan­ning prac­tice. We have estab­lished quan­ti­ta­tive mod­el­ing frame­works to ana­lyze the impacts of AVs on mobil­i­ty sys­tems and urban land use. These frame­works will pro­vide a quan­tifi­able under­stand­ing of the trade­offs, and reveal the under­ly­ing mech­a­nism and iden­ti­fy key para­me­ters that could shape the future of mobil­i­ty sys­tems and urban land use. More­over, the pro­posed mod­el­ing frame­works will aid plan­ning agen­cies with infra­struc­ture adap­ta­tion plan­ning and opti­mize a roadmap for shap­ing high­way infra­struc­ture towards auto­mat­ed mobility.

Behavioral Stability of Cooperative Vehicle Platooning for Energy Saving

Coop­er­a­tive vehi­cle pla­toon­ing is a set of vehi­cles dri­ving togeth­er with low head­way enabled by the con­nect­ed and auto­mat­ed vehi­cle tech­nol­o­gy. Vehi­cle pla­toon­ing has been demon­strat­ed to be a promis­ing way to reduce fuel con­sump­tion and traf­fic emis­sions. Pre­vi­ous stud­ies have pri­mar­i­ly inves­ti­gat­ed opti­mal con­trol of vehi­cle pla­toons to guar­an­tee string sta­bil­i­ty and main­tain the desired space. Few has con­sid­ered the behav­ioral side of the prob­lem, i.e., ensur­ing dri­vers or own­ers’ will­ing­ness to form and main­tain the pla­toon, which is par­tic­u­lar­ly crit­i­cal for human-dri­ven con­nect­ed vehi­cles or pri­vate­ly-owned auto­mat­ed vehi­cles. Because vehi­cles in a pla­toon may ben­e­fit dif­fer­ent­ly from pla­toon­ing, e.g., the lead vehi­cle of the pla­toon may not save any ener­gy at all, some dri­vers or own­ers will not be will­ing to join or stay in a pla­toon even if they are advised to do so. To address this issue, fund­ed by Nation­al Sci­ence Foun­da­tion, we seek for eco­nom­ic mech­a­nisms based on a fair redis­tri­b­u­tion of ben­e­fits to incen­tivize dri­vers or own­ers to form and more impor­tant­ly, main­tain both the sta­bil­i­ty of the “green­est” coop­er­a­tive vehi­cle pla­toon for­ma­tions. The research chal­lenge to be addressed is to achieve, in a dis­trib­uted or local­ized fash­ion, both the inter- and intra-pla­toon sta­bil­i­ty in a dynam­ic and uncer­tain environment.


Modeling and Analysis of Ride-Sourcing Markets

Since their advent in 2009, on-demand ride-sourc­ing com­pa­nies (also called as Trans­porta­tion Net­work Com­pa­nies) such as Uber and Lyft have enjoyed huge suc­cess, but have also cre­at­ed many con­tro­ver­sies. The reg­u­lar cab ser­vices are usu­al­ly reg­u­lat­ed in terms of price, entry and ser­vice qual­i­ty while com­par­a­tive­ly few­er reg­u­la­to­ry require­ments have been imposed for ride-sourc­ing com­pa­nies. Unfair com­pe­ti­tion is argued par­tic­u­lar­ly by cab dri­vers and their employ­ers, who have orga­nized strikes and filed law­suits around the world. Gov­ern­ment offi­cials and leg­is­la­tors are won­der­ing what to do with these ride-sourc­ing com­pa­nies. Their suc­cess has cast­ed doubts on the reg­u­la­tion of the taxi indus­try and chal­lenged some of its fundamentals.

Fund­ed by Nation­al Sci­ence Foun­da­tion, we devel­op method­olo­gies and tools for ana­lyz­ing the struc­ture and com­pe­ti­tion of taxi mar­kets with emerg­ing ride-sourc­ing ser­vices and then deriv­ing insights on their reg­u­la­tion. The research is con­duct­ed in two prin­ci­pal thrusts. The first aims to devel­op ana­lyt­i­cal method­olo­gies to inves­ti­gate opti­mal reg­u­la­tion regimes of taxi mar­kets with ride-sourc­ing com­pa­nies under sim­pli­fied aggre­gate or macro­scop­ic set­tings. The sec­ond devel­ops an agent-based sim­u­la­tion tool and a sim­u­lat­ed test bed to pro­vide crit­i­cal inputs to the the­o­ret­i­cal inves­ti­ga­tions, ver­i­fy the pred­i­ca­tions and insights drawn from the the­o­ret­i­cal tools, and more impor­tant­ly, exper­i­ment and inves­ti­gate var­i­ous reg­u­la­tions and poli­cies in com­plex sys­tems beyond the ana­lyt­i­cal­ly tractable cas­es. This work offers the first com­pre­hen­sive study of taxi sys­tems with emerg­ing ride-source ser­vices and pro­vides the­o­ret­i­cal under­pin­ning for their reg­u­la­tion and management.


Understanding Labor Supply in the Ride-Sourcing Market

As a typ­i­cal exam­ple of a two-sided mar­ket, ride-sourc­ing com­pa­nies serve as inter­me­di­ary plat­forms that effi­cient­ly match cus­tomers to affil­i­at­ed dri­vers near­by. Unlike tra­di­tion­al firm-employ­ee rela­tion­ship, the ride-sourc­ing mar­ket is fea­tured by flex­i­ble labor sup­ply. Dri­vers are free to deter­mine where, when and how long they want to work. Some of them may work full time like pro­fes­sion­al taxi dri­vers while oth­ers only pro­vide ser­vice for very lim­it­ed time. How­ev­er, the flex­i­bil­i­ty enjoyed by ride-sourc­ing dri­vers pos­es chal­lenges for ride-sourc­ing com­pa­nies to man­age the sys­tem. The com­pa­nies may fail to attract suf­fi­cient labor sup­ply despite being able to accu­rate­ly pre­dict days or areas with demand surges. There­fore, there is an increas­ing need to under­stand, esti­mate and main­tain prop­er labor sup­ply in the ride-sourc­ing market.

Fund­ed by Didi Chux­ing, we mod­el dri­vers’ work sched­ul­ing and zon­al choic­es respec­tive­ly, and inves­ti­gate how these deci­sions inter­act with the wage rates that dri­vers receive. Lever­ag­ing empir­i­cal data and evi­dences from Didi Chux­ing, we then apply our mod­els to design and eval­u­ate mech­a­nisms and strate­gies that aim to guide or incen­tivize dri­vers to make bet­ter deci­sions and improve the per­for­mance of the ride-sourc­ing market.

Integrated Operations of Road and Power Networks Coupled by Plug-in Electric Vehicles

Deploy­ment of plug-in elec­tric vehi­cles (PEVs) will lead to more fre­quent and pro­found inter­ac­tion between trans­porta­tion and pow­er sys­tems. The poli­cies and mea­sures imple­ment­ed in the trans­porta­tion sys­tem will change the spa­tial and tem­po­ral dis­tri­b­u­tions of PEVs and thus the pat­tern of their ener­gy require­ment, there­by affect­ing the oper­a­tions of the pow­er sys­tem. On the oth­er hand, the pro­vi­sion of the charg­ing infra­struc­tures and the asso­ci­at­ed charg­ing prices will affect the trav­el pat­terns of PEVs and con­se­quent­ly the oper­a­tions of the trans­porta­tion sys­tem. The lev­el of inter­ac­tion of these two sys­tems will large­ly depend on the mar­ket pen­e­tra­tion of PEVs, advance­ment in charg­ing tech­nolo­gies and, more impor­tant­ly, inno­v­a­tive con­trol strate­gies that lever­age advanced charg­ing tech­nolo­gies to fos­ter the inte­gra­tion of these two systems.

LIMOS has devel­oped mod­els and solu­tion method­olo­gies for ana­lyz­ing the inter­ac­tion between trav­el pat­terns of PEVs and prices of elec­tric­i­ty and roads, design­ing pric­ing poli­cies of both the road and pow­er net­works cou­pled by PEVs, and devel­op­ing deploy­ment strate­gies of pub­lic charg­ing infra­struc­ture to max­i­mize social wel­fare. Our research will enable PEVs to cou­ple the trans­porta­tion and pow­er sys­tems in ways that address crit­i­cal issues in both. The research also advances our under­stand­ing of the com­plex rela­tion­ships, depen­den­cies and inter­de­pen­den­cies that cross the bound­aries of these two sys­tems, and improves the state of the art in trans­porta­tion and pow­er engi­neer­ing, par­tic­u­lar­ly, the the­o­ry of pric­ing of elec­tric­i­ty and roads.


Evaluation of Charging-While-Driving Technology

The deploy­ment of pub­lic charg­ing infra­struc­ture plays a crit­i­cal role in nur­tur­ing the elec­tric vehi­cle (EV) mar­ket and pro­mot­ing the adop­tion of EVs.  Among var­i­ous types of charg­ing tech­nolo­gies, charg­ing-while-dri­ving holds great promise. It can be achieved by either con­duc­tive or induc­tive charg­ing. The for­mer is sim­i­lar to the tech­nol­o­gy used for trams and trains, charg­ing EVs via lines over­head or met­al bars in the pave­ment. The lat­ter, often referred to as dynam­ic wire­less charg­ing, trans­mits pow­er with­out using any phys­i­cal con­nec­tor. The enabling mech­a­nism includes, among oth­ers, induc­tive cou­pling, mag­net­ic res­o­nance cou­pling and microwaves. Giv­en the charg­ing-while-dri­ving tech­nol­o­gy, roads can be elec­tri­fied as charg­ing infra­struc­ture, and EVs can be recharged while they are mov­ing on the charg­ing lanes. Con­se­quent­ly, EV dri­vers may not fear any more run­ning out of bat­tery when they are on the move. Such a per­va­sive wire­less charg­ing plat­form can mit­i­gate or even elim­i­nate the “range anx­i­ety” of EV dri­vers and fur­ther boost the adop­tion of EVs.

Antic­i­pat­ing that charg­ing lanes can be tech­ni­cal­ly ready for deploy­ment in the fore­see­able future, but may be cost­ly to deploy, we estab­lished ana­lyt­i­cal mod­els to inves­ti­gate the eco­nom­ic fea­si­bil­i­ty of charg­ing lanes in sup­port­ing pri­vate and pub­lic trans­porta­tion. Fur­ther­more, math­e­mat­i­cal frame­works have been devel­oped to opti­mize the deploy­ment of charg­ing lanes over trans­porta­tion net­works. Our research offers some guid­ance for the future devel­op­ment of charg­ing-while-dri­ving tech­nolo­gies, and pro­vides a bet­ter under­stand­ing to deci­sion mak­ers on the eco­nom­ic via­bil­i­ty and deploy­ment plan of charg­ing lanes.


In addi­tion, as the word cloud of LIMOS’ past pub­li­ca­tions sug­gests, we have worked exten­sive­ly in devel­op­ing mar­ket-based instru­ments for con­ges­tion mit­i­ga­tion. Below is a sum­ma­ry of our work in this area.

Innovative Market-based Instruments for Congestion Mitigation

Traf­fic con­ges­tion con­tin­ues to threat­en eco­nom­ic pros­per­i­ty and qual­i­ty of life around the world. It is wide­ly acknowl­edged that mar­ket-based instru­ments are an indis­pens­able ingre­di­ent in the recipe for solv­ing the traf­fic con­ges­tion puz­zle, and like­ly to be one of the more effec­tive and cost-effi­cient if prop­er­ly imple­ment­ed. The mar­ket-based instru­ments can be gen­er­al­ly clas­si­fied into two class­es, i.e., price-based and quan­ti­ty-based. The for­mer, wide­ly known as con­ges­tion pric­ing, has been the focus in the trans­porta­tion field for over 90 years. Fund­ed by Nation­al Sci­ence Foun­da­tion, Nation­al Coop­er­a­tive High­way Research Pro­gram and Flori­da Depart­ment of Trans­porta­tion, with a prop­er blend of the­o­ry and prac­tice, we have devel­oped mod­els and algo­rithms to make con­ges­tion pric­ing more prag­mat­ic, flex­i­ble and pub­licly accept­able. Our con­tri­bu­tion to the lit­er­a­ture of con­ges­tion pric­ing includes Pare­to-improv­ing pric­ing to reduce traf­fic con­ges­tion with­out mak­ing any­one worse off; dif­fer­en­ti­at­ed pric­ing of trav­el­ers with dif­fer­ent ori­gins, des­ti­na­tions, or paths; behav­ioral­ly-con­sis­tent pric­ing mod­els that cap­ture bound­ed­ly ratio­nal trav­el behav­iors and a self-learn­ing frame­work for dynam­ic pric­ing of man­aged or high-occu­pan­cy/­tolls (HOT) lanes.

On the oth­er hand, we have inves­ti­gat­ed the quan­ti­ty-based approach, which seeks to direct­ly reg­u­late quan­ti­ty, e.g., trav­el demand. Of par­tic­u­lar inter­est is the so-called cap-and-trade or trad­able cred­it scheme, which cou­ples quan­ti­ty restric­tion with a trad­ing mech­a­nism. Fund­ed by Nation­al Sci­ence Foun­da­tion, our research has pro­vid­ed a bet­ter under­stand­ing of the work­ing mech­a­nism of the quan­ti­ty-based approach to urban con­ges­tion man­age­ment and cre­at­ed a the­o­ret­i­cal frame­work to ana­lyze and design new quan­ti­ty-based schemes.