Optimising Public Transport: Disruptions Included

2019-06-24 Over the past 20 ye­ars, the pu­blic trans­port sys­tem has grown stea­di­ly. Bu­ses in par­ti­cu­lar are of spe­cial re­le­van­ce be­cau­se they tra­vel both in ci­ties and in rural are­as. Sci­en­tists from Leu­pha­na Uni­ver­si­ty and FU Ber­lin now work on a new re­se­arch pro­ject on ro­bust op­ti­mi­sa­ti­on in pu­blic trans­port. The Ger­man Re­se­arch Foun­da­ti­on (DFG) sup­ports the co­ope­ra­ti­on pro­ject with FU Ber­lin, in­iti­al­ly de­si­gned for an 18-month pe­ri­od, with around 280,000 eu­ros.

©Leuphana / Kersten Benecke
Prof. Dr. Lin Xie

Over the past 20 ye­ars, the pu­blic trans­port sys­tem has grown stea­di­ly. Bu­ses in par­ti­cu­lar are of spe­cial re­le­van­ce be­cau­se they tra­vel both in ci­ties and in rural are­as. Sci­en­tists from Leu­pha­na Uni­ver­si­ty and FU Ber­lin now work on a new re­se­arch pro­ject on ro­bust op­ti­mi­sa­ti­on in pu­blic trans­port. The Ger­man Re­se­arch Foun­da­ti­on (DFG) sup­ports the co­ope­ra­ti­on pro­ject with FU Ber­lin, in­iti­al­ly de­si­gned for an 18-month pe­ri­od, with around 280,000 eu­ros.

When Lin Xie has to wait for a bus at a bus stop, she usual­ly thinks of her re­se­arch. The as­sis­tant pro­fes­sor of busi­ness in­for­ma­ti­on sys­tems, in par­ti­cu­lar ope­ra­ti­on re­se­arch, knows how com­plex the lo­gis­tics of trans­port com­pa­nies is. For one thing, long-term plan­ning must be in place: Which bus rou­tes are co­ve­r­ed? How many ve­hi­cles are in ope­ra­ti­on? Which ti­me­ta­ble ap­p­lies? De­s­pi­te a well-plan­ned sys­tem, bus com­pa­nies face new chal­len­ges every day: Traf­fic jams, sick lea­ve of dri­vers, or tech­ni­cal pro­blems are just a few of the many va­ria­bles that bus com­pa­nies have to reckon with at any time. But de­s­pi­te dis­rup­ti­ons, ope­ra­ti­ons must con­ti­nue, and ope­ra­ting costs must not in­crea­se si­gni­fi­cant­ly. Due to the com­ple­xi­ty of traf­fic pro­blems, Lin Xie, in co­ope­ra­ti­on with FU Ber­lin, de­ve­lops pro­to­ty­pe soft­ware that will la­ter be tested by bus com­pa­nies. The soft­ware can be fed with data from past or cur­rent dis­rup­ti­ons: Whe­re is the risk of traf­fic jams par­ti­cu­lar­ly high? Which tech­ni­cal de­fects oc­cur in ve­hi­cles? On which rou­tes do de­lays oc­cur re­pea­ted­ly? The soft­ware is in­ten­ded to cal­cu­la­te so­lu­ti­on mo­dels that keep the con­se­quen­ces of dis­rup­ti­ons as small as pos­si­ble.

The re­se­arch pro­ject “Ro­bust in­te­gra­ted ve­hi­cle-crew-sche­du­ling and crew ros­te­ring in pu­blic tran­sit” fo­cu­ses on the in­te­gra­ted op­ti­mi­sa­ti­on of dai­ly ve­hi­cle sche­du­ling and month­ly crew sche­du­ling with re­gard to dis­tur­ban­ces. Howe­ver, neit­her the in­te­gra­ti­on as­pects nor the ro­bust­ness have been suf­fi­ci­ent­ly re­se­ar­ched so far. Sci­en­tists from Leu­pha­na and FU Ber­lin are now in­vol­ved in the re­se­arch pro­ject to chan­ge this. The first mee­ting of the pro­ject was on June 14 at FU Ber­lin. The aim of the re­se­arch pro­ject is to de­ve­lop a plan­ning mo­del that is cost-ef­fec­tive for com­pa­nies and re­sour­ce-ef­fi­ci­ent on the one hand, and ro­bust against dis­rup­ti­ons on the other. "The best way to do this is to in­clu­de dis­rup­ti­ons in our plan­ning as well. That's why we in­cor­po­ra­te time buf­fers into the sys­tem, but they also cost mo­ney," ex­plains Lin Xie.

For the first time, re­se­ar­chers do not only take into ac­count the tech­ni­cal pro­ces­ses, but also the hu­man com­po­nent: Bus dri­ver sa­tis­fac­tion is ano­ther va­ria­ble in the mo­del­ling pro­cess. "If em­ployees en­joy their work, the bus com­pa­nies also be­ne­fit. Em­ployees who feel their nee­ds are ta­ken se­rious­ly tend to be more re­lia­ble,” ex­plains Lin Xie.