Bisaschi, Luca; Calde oni, Paolo; Ga ces, Inazio; Lecha doy, Lucie; Na doni, Sa a
Wo king Pape
Algo i hmic managemen in he logis ics sec o in F ance
JRC Wo king Pape s Se ies on Labou , Educa ion and Technology, No. 2025/04
P o ided in Coope a ion wi h:
Join Resea ch Cen e (JRC), Eu opean Commission
Sugges ed Ci a ion: Bisaschi, Luca; Calde oni, Paolo; Ga ces, Inazio; Lecha doy, Lucie; Na doni, Sa a
(2025) : Algo i hmic managemen in he logis ics sec o in F ance, JRC Wo king Pape s Se ies on
Labou , Educa ion and Technology, No. 2025/04, Eu opean Commission, Join Resea ch Cen e
(JRC), Se ille
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/322091
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by/4.0/
Cen e
Algo i hmic Managemen in
he Logis ics Sec o in F ance
JRC Wo king Pape s Se ies on
Labou , Educa ion and Technology
2025/04
L. Bisaschi, P. Calde oni, I. Ga ces,
L. Lecha doy, S. Na doni
This publica ion is pa o a wo king pape se ies on Labou , Educa ion and Technology by he Join Resea ch
Cen e (JRC). The JRC is he Eu opean Commission’s science and knowledge se ice.
I aims o p o ide e idence-based scien i ic suppo o he Eu opean policymaking p ocess. The con en s o
his publica ion do no necessa ily e lec he posi ion o opinion o he Eu opean Commission. Nei he he
Eu opean Commission no any pe son ac ing on behal o he Commission is esponsible o he use ha
migh be made o his publica ion. Fo in o ma ion on he me hodology and quali y unde lying he da a used
in his publica ion o which he sou ce is nei he Eu os a no o he Commission se ices, use s should
con ac he e e enced sou ce. The designa ions employed and he p esen a ion o ma e ial on he maps do
no imply he exp ession o any opinion wha soe e on he pa o he Eu opean Union conce ning he legal
s a us o any coun y, e i o y, ci y o a ea o o i s au ho i ies, o conce ning he delimi a ion o i s on ie s
o bounda ies.
Con ac in o ma ion
Name: Ignacio González Vázquez
Add ess: Join Resea ch Cen e, Eu opean Commission (Se ille, Spain)
Email: ignacio.gonzalez- azquez @ec.eu opa.eu
EU Science Hub
h ps://join - esea ch-cen e.ec.eu opa.eu
JRC142292
Se ille: Eu opean Commission, 2025
© Eu opean Union, 2025
The euse policy o he Eu opean Commission is implemen ed by he Commission Decision 2011/833/EU o
12 Decembe 2011 on he euse o Commission documen s (OJ L 330, 14.12.2011, p. 39). Excep o he wise
no ed, he euse o his documen is au ho ised unde he C ea i e Commons A ibu ion 4.0 In e na ional (CC
BY 4.0) licence (h ps://c ea i ecommons.o g/licenses/by/4.0/). This means ha euse is allowed p o ided
app op ia e c edi is gi en and any changes a e indica ed. Fo any use o ep oduc ion o pho os o o he
ma e ial ha is no owned by he EU, pe mission mus be sough di ec ly om he copy igh holde s.
All con en © Eu opean Union 2025
How o ci e his epo : Bisaschi, L., Calde oni, P., Ga ces, I., Lecha doy, L. and Na doni, S., Algo i hmic
Managemen in he Logis ic Sec o in F ance, Eu opean Commission, Se ille, 2025, JRC142292.
1
Con en s
Abs ac ..................................................................................................................................................................................... 2
Execu i e summa y ............................................................................................................................................................ 3
1 In oduc ion ................................................................................................................................................................... 4
2 Me hodology ................................................................................................................................................................. 5
3 Applica ions o algo i hmic managemen ools in he logis ics sec o in F ance and i s
possible impac s ................................................................................................................................................................... 6
3.1 F ench s a egy o A i icial In elligence ....................................................................................... 6
3.2 Algo i hmic managemen ools in he logis ics sec o ........................................................... 7
4 Business model and wo k o ganisa ion ........................................................................................................ 9
4.1 Business model and he deli e y o se ices ............................................................................... 9
4.2 O ganisa ion and coo dina ion o wo k p ocesses ................................................................. 14
4.3 Occupa ion, asks and skills ................................................................................................................ 15
5 Wo king condi ions ................................................................................................................................................. 16
5.1 Wo k in ensi ica ion ................................................................................................................................. 16
5.2 Au onomy ...................................................................................................................................................... 17
5.3 Social in e ac ion ....................................................................................................................................... 17
6 E hics o AI .................................................................................................................................................................. 17
7 Su eillance and da a p i acy issues in he e a o algo i hmic managemen .................... 18
8 Conclusions ................................................................................................................................................................. 24
Re e ences ............................................................................................................................................................................ 26
Lis o igu es...................................................................................................................................................................... 28
Lis o boxes ........................................................................................................................................................................ 29
Appendix ................................................................................................................................................................................ 30
1. Dis ibu ion o in e iews ........................................................................................................................... 30
2
Abs ac
This wo king pape p o ides an assessmen o he impac s o algo i hmic managemen (AM) ools
on wo king condi ions, job quali y, and indus ial ela ions in he logis ics sec o o F ance. I
analyses he di e en ways algo i hmic managemen echnologies a ec coo dina ion, dis ibu ion
and con en o asks, employee au onomy, job p o iles, and indus ial ela ions in logis ics
companies. The epo p esen s a comp ehensi e s udy conduc ed h ough desk esea ch,
s akeholde consul a ion, and in e iews wi h F ench s a -ups, indus y associa ions, public
egula o y bodies, academia, and la ge companies. The F ench S a egy o A i icial In elligence
and he ecen expansion o AI in F ench companies a e discussed, along wi h some key
applica ions o algo i hmic ools. While algo i hmic managemen ools o e se e al bene i s o
logis ics companies, including op imizing ope a ions and enhancing cus ome sa is ac ion, he e a e
also conce ns abou he e hical and legal implica ions o ising echnologies, pa icula ly ega ding
wo ke p i acy and he po en ial o bias o disc imina ion. Recen egula ions and guidelines a e
p esen ed o ensu e he ai and anspa en use o hese ools. The pape concludes by discussing
he need o balance he bene i s o algo i hmic ools wi h he po en ial isks and e hical conce ns,
and he ole o he ma ke and public policy in he coo dina ion o economic ac i i ies in he
logis ics sec o .
Keywo ds: Algo i hmic Managemen , Logis ics, Wo king Condi ions, Job Quali y
Join Resea ch Cen e e e ence numbe : JRC142292
3
Execu i e summa y
The epo p esen s a selec ion o case s udies o algo i hmic managemen echnologies (AMT) in
he logis ics sec o in F ance. Speci ically, he case s udies examine he use o a ious AMTs,
including AI ools, which a e inc easingly deployed o manage wo ke s in he sec o .
F ance is home o some o he wo ld's majo logis ics playe s, including shipping companies, pa cel
deli e ies, eigh s deli e ies, o e minal ope a o s, and many o he la ges logis ics companies in
he wo ld ope a e on F ench soil. F ench go e nmen s ha e ied o os e he de elopmen o new
s a -ups, some o which ope a e in he de elopmen o AMTs and AI echnologies o manage
physical and in o ma ional lows. The logis ics indus y is a signi ican con ibu o o he F ench
economy, accoun ing o 10% o GDP, 150,000 i ms, and 1.8 million jobs. The F ench go e nmen
has in es ed o e 1.5 billion eu os in AI de elopmen , and p i a e co- inancing has added an
addi ional 500 million eu os.
While algo i hmic managemen ools ha e bene i ed he logis ics indus y, i has also c ea ed
ques ions abou e hical applica ions and echnological use in da a p o ec ion and p i acy.
In e iewees explained ha algo i hms ha e enabled machines o sense, comp ehend, lea n, and
ac a human-like le els, aising conce ns abou hei implica ions o employees, such as loss o
au onomy, inc eased su eillance, and biased decision-making. Mo eo e , wo ke s, especially
pla o m wo ke s, ha e mobilised agains algo i hmic managemen , con es ing i s impac on hei
wo king condi ions, as F ance has his o ically been a land o wo ke s' mobilisa ion in he logis ics
sec o . Fo example, F ench ci il socie y has engaged wi h non-p o i expe imen s in he u ban
biking deli e y sec o , and ade unions ha e opened a deba e on he impac o new digi al
echnologies on wo king condi ions.
The s akeholde consul a ion p esen ed in his epo sugges s ha he inc easing adop ion o AMTs
in he logis ics sec o in F ance has bo h posi i e and nega i e impac s. While i has enabled
companies o au oma e and s eamline p ocesses, p edic demand, and educe ope a ional cos s, i
has also c ea ed ques ions abou e hical applica ions and echnological use in da a p o ec ion and
p i acy. This has aised conce ns abou AMTs’ impac on wo ke s' au onomy, inc eased su eillance,
and biased decision-making. The e o e, s akeholde s should conside he implica ions o AMT on
wo king condi ions in he logis ics sec o , and design measu es o mi iga e he nega i e impac s o
algo i hmic managemen .
Main indings
The indings e eal ha AMT adop ion has become widesp ead in he F ench logis ics indus y in
ecen yea s, o e ing compe i i e ad an ages o companies ha in es in i . Many logis ics
companies ha e in es ed in AMT o au oma e and s eamline p ocesses and o p edic and o ecas
demand, esul ing in lowe ope a ional cos s and highe e enues. Fo example, Sma Wa ehouse
Sys ems a e used o ecognise pa e ns and dependencies om uns uc u ed da a using IoT, AI, and
cloud compu ing. These appliances can adap independen ly and dynamically o new ci cums ances
h oughou he en i e logis ics sys em, making mono onous jobs simple , and ope a ions mo e
e icien and cos -e ec i e. Howe e , his has also made ce ain wa ehouse jobs edundan and
inc eased he impo ance o enginee ing and manage ial jobs.
Quick guide
The i s pa o his epo in oduces he opic and i s ele ance. Sec ion 2 desc ibes he
me hodology ollowed h oughou he s udy, while Sec ion 3 p esen s a desc ip ion o he logis ics
sec o in F ance and he echnologies used he e. Sec ion 4 lays he g oundwo k o he e idence
collec ed h oughou he s akeholde consul a ions wi h i s -hand da a and esponses on he
di e en impac s o algo i hmic managemen ools on business model and he deli e y o se ices,
Sec ion 5 on wo king condi ions, Sec ion 6 on he e hics o AI, and Sec ion 7 on su eillance and
da a p i acy issues. Finally, Sec ion 8 concludes he epo wi h a ecapi ula ion o he da a
collec ed and i s di e en impac s.
4
1 In oduc ion
The apid ad ancemen o he digi al e olu ion and he g owing connec i i y in he wo kplace a e
eshaping he wo ld o wo k. Digi al ools and algo i hmic managemen p ac ices a e now being
adop ed in adi ional wo kplaces as mechanisms o wo k coo dina ion. The digi alisa ion o wo k,
along wi h digi al moni o ing and algo i hmic managemen – de ined as he use o compu e -
p og ammed p ocesses o coo dina e labou wi hin an o ganisa ion – c ea es new business
oppo uni ies, enhances e iciency, and op imizes wo k lows. Howe e , his shi also aises
conce ns abou wo king condi ions, he po en ial decline in job quali y, and he inc eased isk o
wo ke su eillance
1
.
O e he pas ew decades, he use o algo i hms ac oss indus ies has quickly sp ead h oughou
he wo ld. Especially impac ing he mos indus ialised coun ies, algo i hms ha e ans o med ou
socie ies, how companies and ma ke s ope a e, and how humans in e ac wi h machines: om
academia o business p ac ices, and om ope a ions o human managemen . Algo i hmic
managemen e e s o he use o AI and o he algo i hmic echniques o manage and supe ise
wo ke s and business p ocesses. Essen ially, i in ol es au oma ing ce ain manage ial asks ha
would no mally be ca ied ou by human manage s. In p ac ice, algo i hmic managemen
echnologies (AMT) can ake many di e en o ms. Fo example, hey can in ol e using machine
lea ning (ML) algo i hms o analyse employee pe o mance da a and make decisions abou
p omo ions, bonuses, and o he o ms o compensa ion. They can also in ol e using cha bo s o
o he au oma ed ools o communica e wi h employees and answe hei ques ions o using
p edic i e analy ics o o ecas u u e business needs and adjus s a ing le els acco dingly.
Digi al labou pla o ms ha e been pionee s in adop ing su eillance echnologies and algo i hmic
managemen p ac ices
2
,
3
. This has been acili a ed bo h by hei inhe en ly digi al ope a ing
en i onmen and by he in o mal na u e o hei wo king ela ionships, which has o en allowed
hem o bypass many o he es ic ions ypically ound in adi ional wo kplaces. Howe e ,
algo i hmic managemen and digi al su eillance a e now inc easingly pe mea ing con en ional
wo k se ings as well, p omp ing b oade conce ns abou hei impac on labou p ocesses, he
o ganisa ion o wo k, and he balance o powe in he wo kplace
4
. Fo example, he e a e conce ns
abou algo i hmic bias, whe e algo i hms may inad e en ly disc imina e agains ce ain g oups o
wo ke s o pe pe ua e exis ing inequali ies. The e a e also conce ns abou wo ke p i acy and
au onomy, as well as po en ial legal liabili y i algo i hmic decisions lead o nega i e ou comes o
wo ke s. The e o e, i is impo an o o ganisa ions o ca e ully conside he implica ions o
algo i hmic managemen and ensu e ha hey ha e obus policies and p ocedu es in place o
add ess hese issues.
The case s udies p esen ed in his epo we e conduc ed a he end o 2022 in he ame o a
b oade esea ch s udy commissioned by he Join Resea ch Cen e (JRC) o Open E idence,
in es iga ing he impac o algo i hmic managemen ools (AMT) on wo k o ganisa ion. The
esea ch s udy was pa o a b oade p ojec ca ied ou join ly by he JRC and he In e na ional
Labou O ganisa ion (ILO) ocusing on coun y-sec o analyses in he logis ics and heal hca e
1
Rani, U., Pesole, A. and Gonzalez Vazquez, I., Algo i hmic Managemen p ac ices in egula wo kplaces: case s udies in
logis ics and heal hca e, Publica ions O ice o he Eu opean Union, Luxembou g, 2024, doi:10.2760/712475, JRC136063.
2
Pesole, A., U zì B anca i, M.C., Fe nandez Macias, E., Biagi, F. and Gonzalez Vazquez, I. 2018. Pla o m Wo ke s in Eu ope
E idence om he COLLEEM Su ey, EUR 29275 EN, Publica ions O ice o he Eu opean Union, Luxembou g, 2018, ISBN
978-92-79-87996-8, doi:10.2760/742789, JRC112157.
3
ILO. 2021. Wo ld Employmen and Social Ou look 2021: The Role o Digi al Labou Pla o ms in T ans o ming he Wo ld
o Wo k. In e na ional Labou O ganiza ion.
4
Baiocco, S., Fe nández-Macías, E., Rani, U. and Pesole, A., 2022. The Algo i hmic Managemen o wo k and i s implica ions
in di e en con ex s, Se ille: Eu opean Commission, JRC129749.
5
sec o s in wo Eu opean coun ies (F ance and I aly) and wo non-EU coun ies (India and Sou h
A ica) allowing o a compa a i e analysis ac oss coun ies wi h di e en le els o de elopmen
5
.
The aim o he o e all p ojec was o assess applica ions o algo i hms in he wo kplace and
examine hei impac s on he coo dina ion o wo k and wo king condi ions, oge he wi h he ole o
social dialogue in he adop ion and implemen a ion o hese echnologies. The logis ics sec o was
chosen as an example o a highly digi alised sec o in which he use o digi al moni o ing and
algo i hmic managemen echnologies has been widely documen ed in he li e a u e. In con as , in
he heal hca e sec o he use o algo i hmic managemen as pa o digi al heal h pla o ms is
mo e ecen and less well documen ed. This epo ocuses on he logis ics sec o in F ance.
2 Me hodology
Each case s udy conside ed in he p ojec was concei ed o include a numbe o quali a i e, in-
dep h, semi-s uc u ed in e iews and ield isi s in an es ablishmen using algo i hmic
managemen o he coo dina ion o wo k p ocesses, o be complemen ed by desk esea ch as
app op ia e
6
. Speci ically, he case s udies in ol ed in e iews wi h indi iduals in di e en unc ions
and oles in each es ablishmen including manage s a di e en le els, union and wo ke
ep esen a i es, echnology specialis s, and wo ke s a ec ed by he echnology. A semi-s uc u ed
in e iew guide was designed and o ganised a ound a se o co e hemes and ques ions in ended o
in es iga e he impac o algo i hms and digi al solu ions on wo k o ganisa ion and wo king
condi ions. All in e iews we e eco ded and la e ansc ibed. A con en analysis o each in e iew
was ca ied ou : in e iews’ ansc ip s we e indi idually s udied and de eloped om gene alised
con en s owa ds obus analyses p og essi ely c ys alising on he key dimensions co e ed. The
main opics co e ed in he in e iews include: (i) impac s on business model and he deli e y o
se ices; (ii) impac s on o ganisa ion and coo dina ion o wo k p ocesses; (iii) impac s on
occupa ions, asks and skills; (i ) impac s on wo k in ensi ica ion; ( ) impac s on au onomy; ( i),
impac s on social in e ac ion; ( ii) e hical conce ns su ounding he use o AI; ( iii) impac s on
moni o ing and su eillance; and (ix) impac s on da a p i acy issues.
The pa icipa ion and in ol emen o di e en s akeholde s in he esea ch allows o p o ide a
balanced unde s anding o he implemen a ion o he digi al solu ion, he le el o algo i hmisa ion
o he echnology in place, and i s impac s in he wo kplace. In addi ion o he quali a i e in e iews,
he case s udies a e supplemen ed wi h in o ma ion and documen a ion p o ided by he
in e iewees and o he wo king pape s.
Ou each ac i i ies o he selec ion o he es ablishmen s o be used as case s udies in he F ench
logis ics sec o in ol ed accessing logis ics companies o all sizes ac oss F ance, including
mul ina ionals and s a -ups. Howe e , om he beginning o he p ojec , F ench logis ics
companies we e pa icula ly eluc an o pa icipa e in he s udy. The main easons o companies
declining pa icipa ion as a case s udy a he es ablishmen -le el included conce ns on he
legisla i e p ocess o he EU, ime cons ain s, lack o in e es , and una ailabili y due o lack o
esou ces. Mo eo e , he opic o algo i hmic managemen in he logis ics sec o has become
pa icula ly sensi i e in F ance due o ecen scandals and ongoing wo ke p o es s. Conce ns o e
p eca ious wo king condi ions, su eillance, and he pe cei ed lack o anspa ency in algo i hmic
decision-making ha e ueled c i icism. As a esul , companies may be eluc an o expose
hemsel es o public sc u iny o egula o y a en ion, making hem hesi an o pa icipa e in he
s udy. In o al, a ound 50 di e en companies using algo i hmic managemen ools we e con ac ed
o pa icipa e as a case s udy a he es ablishmen -le el. Consequen ly, he me hodology o he
esea ch s udy o he F ench logis ics was modi ied o o e come he conce ns aised by companies
and o collec su icien da a whils main aining he o e a ching scope o he esea ch p ojec . In
5
Rani, U., Pesole, A. and Gonzalez Vazquez, I., Algo i hmic Managemen p ac ices in egula wo kplaces: case s udies in
logis ics and heal hca e, Publica ions O ice o he Eu opean Union, Luxembou g, 2024, doi:10.2760/712475, JRC136063.
6
Ibid.
6
he end, a o al o 25 s akeholde s om di e se backg ounds in logis ics we e in e iewed ac oss
14 di e en o ganisa ions o en e p ises. This enabled o gain a be e unde s anding o he impac
o algo i hmic managemen ools in he wo kplace. S akeholde s included F ench s a -ups and
la ge logis ics companies, indus y associa ions, academia and public egula o y bodies. All
in e iews, desk esea ch, da a collec ion and epo ing we e ca ied ou be ween Decembe and
Ma ch 2023. In e iews we e conduc ed online ia MS Teams wi h a p e-es ablished ques ionnai e
sen in ad ance (see appendix o mo e de ails).
3 Applica ions o algo i hmic managemen ools in he logis ics
sec o in F ance and i s possible impac s
3.1 F ench s a egy o A i icial In elligence
Acco ding o he F ench go e nmen , he logis ics indus y in F ance accoun s o 10% o he GDP,
150,000 i ms, and 1,8 million jobs ( ou imes as much as he au omobile indus y)
7
. Mo eo e ,
be ween he yea s 2018-2023, he F ench go e nmen has dedica ed mo e han 1.5 billion eu os o
he de elopmen o AI in he coun y
8
, and an addi ional 500 million eu os coming om p i a e co-
inancing has been pou ed in o he ech indus y
9
.
The F ench Na ional S a egy on AI (S a égie na ionale pou l'in elligence a i icielle), labelled ‘AI
o Humani y’, was launched in 2018 and aims a implemen ing he ecommenda ions o Céd ic
Villani’s epo on AI launched ha same yea (Villani e al., 2018). The epo builds on ou axes
10
.
Axis 1: To enhance he F ench esea ch ecosys em on AI h ough addi ional inancial esou ces,
new in as uc u es, and mo e links wi h he indus y. Axis 2: To accele a e AI dissemina ion in he
economy, public adminis a ion, and socie y. Axis 3: To encou age he de elopmen o a
‘T us wo hy AI’, bo h a he na ional and in e na ional le els. Axis 4: To de elop and ein o ce AI
educa ional p og ammes o inc ease AI con en ac oss uni e si ies. All s akeholde s in e iewed
ag eed ha he Na ional S a egy has played an impo an ole in he adop ion o new AI ools in
he indus y, as well as algo i hmic managemen echnologies ac oss di e en sec o s. Th ough he
aid o s a e inancing and he pooling oge he o alen , in e iewees ag eed ha a new push in
he use o AI in small and medium en e p ises was g owing, including new AI ools being in en ed in
F ance.
The Villani epo highligh s he need o build a da a- ocused economic policy o ca apul Eu ope
in o becoming a wo ld leade in AI. The epo sugges s ha F ance and Eu ope mus design a
ailo ed model o capi alize on high p o ec ion s anda ds h ough he Eu opean Gene al Da a
P o ec ion Regula ion (GDPR)
11
. To achie e his, public au ho i ies mus in oduce new ways o
p oducing, sha ing and go e ning da a by making da a a common good. The S a e is a key d i e
in hese a ious a eas o ans o ma ion, and public au ho i ies mus ensu e ha hey adop he
necessa y ma e ial and human esou ces o ac o AI in o he way hey add ess public policy, wi h
he aim o bo h pu suing mode niza ion and ac ing as an example o be ollowed (Villani e al.,
2018).
Addi ionally, he epo sugges s se ing up sha ed sec o pla o ms ha p o ide secu e and ailo ed
access o di e en pa icipan s in hese ecosys ems, including esea che s, companies, and public
au ho i ies. Mo eo e , hese sha ed pla o ms would enable sha ing o use ul da a o he
de elopmen o AI, as well as os e esou ces and ex ensi e compu ing in as uc u e (Villani e al.,
7
Minis è e de la T ansi ion écologique e de la Cohésion des e i oi es (2021). h ps://www.ecologie.gou . /logis ique-en
ance#:~: ex =La%20logis ique%20es %20%C3%A0%20ce,%2C8%20million%20d'emplois
8
F ance AI S a egy Repo (2021). h ps://ai-wa ch.ec.eu opa.eu/coun ies/ ance/ ance-ai-s a egy- epo _en
9
F ench AI S a egy (2018). h ps://www.eu ac i .com/sec ion/digi al/news/ ench-ai-s a egy- ech-sec o - o- ecei e-o e -
e2-bln-in-nex -5-yea s/
10
OECD.AI Policy Obse a o y (2021). h ps://oecd.ai/en/dashboa ds/policy-ini ia i es/h p:%2F%2Faipo.oecd.o g%2F2021-
da a-policyIni ia i es-25374
11
AI Fo Humani y (2018). h ps://www.ai o humani y. /
13
— Rou ing and scheduling: Wi h he use o algo i hms, logis ics companies can op imize ou es
and schedules o imp o e deli e y imes and educe cos s. These ools can analyse da a on
a ic pa e ns, wea he condi ions, and deli e y des ina ions o de e mine he mos e icien
ou es.
— In en o y managemen : Algo i hms can moni o in en o y le els in eal- ime and make
ecommenda ions on when and how much o eo de . These ools can help logis ics companies
a oid s ockou s and minimize was e.
— Quali y con ol: Algo i hms can be used o moni o and analyse da a om senso s and o he
sou ces o iden i y quali y con ol issues in eal- ime. This can help logis ics companies iden i y
and add ess p oblems be o e hey lead o cus ome complain s o p oduc ecalls.
— P edic i e main enance: Algo i hms can also analyse da a om senso s and o he sou ces o
p edic when equipmen is likely o ail. These ools help logis ics companies schedule
main enance p oac i ely, educing down ime and ex ending he li espan o equipmen .
— Cus ome se ice: Finally, algo i hms can be used o analyse da a on cus ome p e e ences
and beha iou s as o pe sonalize cus ome se ice in e ac ions. This can in ol e ecommending
p oduc s o se ices, p o iding ailo ed o e s o discoun s, and esponding o cus ome
inqui ies in eal- ime.
Box 2. Case s udy on F ench Logis ics 3
A majo p og amme wi hin he Founda ion is a las -mile logis ics pla o m o he op imisa ion o
deli e y lows, a ic p edic ion, and mobili y ansi ion o deli e y ehicle lee s. One algo i hmic
managemen ool wi hin pla o m allows managemen o he o ganisa ion o deli e y a eas by
adap ing i s needs and de eloping he oadway acco ding o he equi emen s o d i e s. Thanks o
he ool, d i e s ha e easy access o impo an in o ma ion be o e and h oughou de i ing goods
and when planning u u e ou es. This includes in o ma ion on oad a ic and deli e y a eas, he
numbe and loca ion o deli e y a eas, he a ailabili y o deli e y a eas, he ypology o pa ked
ehicles, he du a ion o pa king zones, and he na u e o pa king a eas. The ool sha es simila i ies
wi h ano he algo i hm used by Amazon as a las -mile solu ion o as e deli e y, lowe cos s, and
a be e cus ome expe ience – quickly and easily managing i s logis ics in a ew clicks
27
.
In essence, he ool de eloped by F ench Logis ics 3 enables digi alising comme cial ehicles’ disc
pa king in F ance (a sys em o allowing ime- es ic ed ee pa king h ough he display o a
pa king disc showing he ime a which he ehicle was pa ked). The i s s ep in using he ool is a
simpli ied pa king decla a ion, wi h in o ma ion on pa king ime, loca ion, and ale s, ia sel -
geoloca ion o deli e y spaces. The second s ep is o moni o and in o m he d i e o know whe e
hey can pa k, wi h geoloca ion o deli e y a eas and he a ailabili y le el o each; b oken down
be ween ee, a ailable o 5mins and una ailable. The hi d s ep is o ‘consul and adap ’, wi h
d i e s being able o ese e pa king spaces and deli e y a eas in ad ance. Figu e 1 illus a es he
ool’s dashboa d. On he le -hand side, we see he mobile in e ace showing he pa king ime limi ,
loca ion, and o he key in o ma ion o d i e s, and on he igh -hand side, we see a hea map wi h
pa king a ailabili y b oken down in o he h ee a o emen ioned le els.
Figu e 1: Example o dashboa d
27
AWS (2023). h ps://aws.amazon.com/blogs/supply-chain/aws-las -mile-solu ion- o - as e -deli e y-lowe -cos s-and-a-
be e -cus ome -expe ience/
14
Sou ce: Ma e ial p o ided by F ench Logis ics 3
The ope a ing p inciples o he ool a e simplici y and con iden iali y, wi h a decla a ion o pa king
unde a simple click, a geoloca ion only a he ime o pa king (inac i e when he ehicle is mo ing
o limi da a p i acy issues), and phone ale s on he emaining pa king ime. Mo eo e , he algo i hm
allows o managing he e olu ion o he d i e ’s needs by simula ing he impac s on a ic. The
algo i hm p edic s a ic and ee pa king spaces using da a om public came as and senso s o
ecognise ca s, mo o bikes, and bicycles on he s ee . By using his ool, companies can cen alise
he di e en deli e y ou es o ehicle lee s and easily manage hei logis ics om one space. Wi h
li e upda es o he deli e y s a us, posi ion and pa king ime, companies can manage he a ailable
esou ces a all imes and be e o ganise hemsel es a ound peak ac i i ies. Using he ool helps
inc ease low op imisa ion and inc ease ou e e iciency.
4.2 O ganisa ion and coo dina ion o wo k p ocesses
Algo i hmic managemen ools can change he way wo k is o ganized and managed. Manage s may
ely mo e hea ily on da a and analy ics o make decisions abou how wo k is assigned and
pe o med. This can esul in wo ke s being assigned asks based on da a-d i en me ics, such as
p oduc i i y o e iciency, a he han hei own p e e ences o skills. Fo example, in a call cen e,
an algo i hmic managemen ool may assign calls o wo ke s based on hei p e ious call
pe o mance, a he han allowing wo ke s o choose which calls hey ake based on hei own
p e e ences o skills.
One example o how algo i hmic managemen ools impac wo k o ganiza ion in he F ench
logis ics sec o is h ough he use o ou e op imiza ion algo i hms o deli e y d i e s. These
algo i hms use eal- ime a ic da a o calcula e he mos e icien ou es o d i e s o ake o
comple e hei deli e ies, educing a el ime and inc easing p oduc i i y.
Ano he example is he use o p edic i e analy ics ools o o ecas demand and op imize in en o y
le els in wa ehouses. This can help o educe o e s ocking and unde s ocking o goods, which can
imp o e e iciency and educe was e.
Addi ionally, he use o algo i hmic managemen ools may equi e new skills and aining o
wo ke s, pa icula ly in he use o da a analy ics and echnology. This can lead o changes in job
oles and esponsibili ies and may equi e addi ional in es men in aining and de elopmen
p og ams.
To collec in o ma ion on he impac s o AMTs on wo k coo dina ion, o ganisa ion and wo k lows,
F ench AI associa ions and non-p o i echnology o ganisa ions we e in e iewed. The co- ounde o
a F ench AI associa ion ounded ollowing he Villani Repo ha helps s a -ups aise unds, sha e
ideas, and pool alen oge he . F ench logis ics associa ion has se e al wo king g oups on AI,
including one on HR ocusing on he eo ganisa ion o eams, he eplacemen o humans, and he
ques ions o eassigning employees. O he wo king g oups in ol e he anspo and deli e y sec o ,
supply chain, s ock managemen , low op imisa ion, and educing dange s in wa ehouses. The
15
ounde explained ha he objec i e is o ha e F ench digi al so e eign y o no depend on o he
coun ies and be able o compe e wi h Silicon Valley.
Fo he ounde , he goal o AI applica ions is gene ally o lowe cos s, dec ease he ime needed o
comple e asks and inc ease he equency, eac i i y, and p o i abili y o eams. Se e al case
s udies wi hin he associa ion include new s anda ds on he espec ul use o d ones in he indus y.
The in e es is based on he inc easing use o d ones in wa ehouse s o age a eas o coo dina e
employees and manage wo ke s in ac o ies. Likewise, d ones a e used o sa e y, low
op imisa ion, and isk educ ion. Ano he example discussed was he inc easing use o d ones o
anspo goods be ween emo e a eas and islands a ound he wo ld. Whils be o e human-pilo ed
helicop e s we e widely employed o anspo essen ial goods in emo e a eas, such as medicine
ac oss sho dis ances and be ween islands, nowadays d ones a e being used o deli e ies. One
example discussed wi h he associa ion was he use o d ones in u al Canada o deli e medical
supplies o emo e a eas and indigenous communi ies
28
. Many echnology expe s ealised ha hey
could au oma e logis ics by sending unmanned d ones ha a e mo e uel e icien , as e , and
mo e en i onmen ally iendly, said he in e iewee. Fu he mo e, he s a ed ha d ones we e being
used ac oss F ench companies o ai li s - loca ion, ai cap u e, ada , came a e c.
Ano he F ench supply chain associa ion in e iewed explained ha AI managemen ools we e
mainly being used in Wa ehouse Managemen Sys ems (WMS) and T anspo Managemen Sys ems
(TMS). Ac oss he anspo and wa ehouse applica ions, digi al echnologies a e apidly expanding
o manage employees and inc ease wo k o ganiza ion, he in e iewee said.
4.3 Occupa ion, asks and skills
Algo i hmic managemen ools can equi e wo ke s o de elop new echnical skills o use and
in e ac wi h he algo i hms. Wo ke s may need o be ained in how o use digi al de ices o how o
in e p e he da a hey gene a e. O en, hese wo asks a e assigned o di e en ca ego ies o
wo ke s: low- ank wo ke s mainly p oduce in o ma ion, while medium-high ank wo ke s mainly
in e p e and manipula e in o ma ion.
28
The Uni e si y o B i ish Columbia (2022). h ps://beyond.ubc.ca/ o wa d-happens-he e- u al-heal h/
16
Figu e 2: Task dis ibu ion and job hie a chisa ion acco ding o ma gins o da a manipula ion a
Amazon.com wa ehouses
Sou ce: Massimo 2020a
In addi ion, wo ke s may need o be com o able wi h echnology and da a analysis in o de o
e ec i ely use he algo i hmic ools. Fo example, a cus ome se ice ep esen a i e may need o
lea n how o use a cha bo o espond o cus ome inqui ies, o how o analyse cus ome da a o
iden i y pa e ns and ends. Some examples o how AMT a e impac ing skills in he F ench
Logis ics sec o include:
— Technical skills: The use o AMT equi es employees o ha e echnical skills o ope a e and
in e p e he da a gene a ed by hese ools. Fo example, wo ke s need o be p o icien in using
wa ehouse managemen sys ems, in en o y managemen so wa e, and o he simila ools o
manage and op imize hei wo k.
— Analy ical skills: The da a gene a ed by AMT p o ides aluable insigh s in o he pe o mance
o logis ics ope a ions. Consequen ly, logis ics wo ke s need o ha e analy ical skills o in e p e
his da a and make in o med decisions based on he insigh s gained.
— Communica ion skills: Algo i hmic managemen ools a e changing he way logis ics wo ke s
communica e wi h each o he and wi h hei supe io s. Fo example, some logis ics companies
a e using cha bo s o alk wi h hei employees and p o ide hem wi h eal- ime upda es on
hei wo k.
— Adap abili y: The use o AMT equi es logis ics wo ke s o be adap able and lexible. They
need o be able o quickly lea n and adap o new ools and p ocesses as hey a e implemen ed.
— Mul i asking: Algo i hmic managemen ools a e helping logis ics wo ke s o become mo e
e icien and p oduc i e by au oma ing ou ine asks. As a esul , wo ke s need o be able o
mul i ask and handle mul iple asks simul aneously o maximize hei p oduc i i y.
5 Wo king condi ions
5.1 Wo k in ensi ica ion
Algo i hmic managemen ools can inc ease wo k in ensi y o logis ics wo ke s, because wo ke s
may be equi ed o comple e asks mo e quickly o wi h g ea e accu acy, as he ool p io i izes
17
speed and e iciency. This can lead o inc eased s ess and bu nou among wo ke s. Mo eo e ,
algo i hmic managemen ools can dis up wo king condi ions o logis ics wo ke s. Fo example,
wo ke s guided by AI-powe ed ou e planning ools may be equi ed o wo k longe hou s o on a
mo e lexible schedule in o de o mee he demands o he ool.
5.2 Au onomy
The use o AMT can educe he au onomy o logis ics wo ke s, since wo ke s may be gi en less
con ol o e hei schedules and asks, as he algo i hm de e mines he mos e icien way o
comple e he wo k. This can happen, o example, when wo k o ce di ec ion and ask alloca ion a e
managed h ough ins uc ions deli e ed ia handheld o wea able digi al de ices. Wo ke s use
scanne s o ansmi eal- ime ope a ional upda es, while hei de ices simul aneously ack me ics
such as ask comple ion speed and e iciency. This da a eeds in o cen al algo i hms ha analyse
a iables like loca ion, mo emen , and iming o s eamline wo k lows and assign asks acco dingly.
As a esul , wo ke s ha e minimal lexibili y o challenge o modi y ins uc ions, as algo i hmic
decisions a e p ede e mined o op imize e iciency
29
. This can also lead o wo ke s eeling less
engaged and less sa is ied wi h hei jobs, and, as a consequence, o a loss o lexibili y o he
o ganiza ion
30
.
5.3 Social in e ac ion
Algo i hmic managemen ools can impac in e ac ions be ween logis ics wo ke s and hei
colleagues, supe iso s, and cus ome s. Fo ins ance, wo ke s may be equi ed o communica e
mo e equen ly and mo e quickly in o de o mee he demands o he ool. On he o he hand,
wo ke s may in e ac less wi h hei colleagues, as hey a e ocused on comple ing hei indi idual
asks as e icien ly as possible. This can lead o a mo e isola ed and impe sonal wo k
en i onmen
31
,
32
.
6 E hics o AI
The e hical issues o algo i hmic managemen ools we e u he discussed wi h a uni e si y
p o esso a Pa is-So bonne Uni e si y specialised in AI and e hics. She a i med ha companies
using AMTs mus in es esou ces in acqui ing expe s in quali y o wo k. This is o ensu e he
ecommended me hodologies o unde s anding e hical isks in a mul i-agen , human-machine
sys em a e me . Once a mo al mapping o he si ua ion is made, his allows hose who build
solu ions o do so 'e hically by design' o con ol a ce ain numbe o e hical cha ac e is ics and
hen moni o coope a ion be ween humans and machines. Ul ima ely, ex e nal s akeholde s mus
ecommend audi s wi h se c i e ia, and humans o check he causes o bias, he p o esso said. We
canno unde es ima e he isks o AI, she said, we ha e o accep hem, bu also be mo e c ea i e.
The EU AI Ac has cha ac e ised he classi ica ion o isks, so we mus no minimise
he isks no in inge on he c ea i i y o companies. We a e wo king on he issue o
‘ us wo hy AI’ and linguis ic nudges. These a e mechanisms ha we desc ibe o
obse e e hical ensions. The digi al gian s used o ha e e hics commi ees, bu hey
a e in he p ocess o i ing hem as pa o hei ecen edundancy plans.
AI and E hics P o esso , Pa is-So bonne Uni e si y
29
Rani, U., Pesole, A. and Gonzalez Vazquez, I., Algo i hmic Managemen p ac ices in egula wo kplaces: case s udies in
logis ics and heal hca e, Publica ions O ice o he Eu opean Union, Luxembou g, 2024, doi:10.2760/712475, JRC136063.
30
F iedman, And ew. 1977. ‘Responsible Au onomy Ve sus Di ec Con ol O e he Labou P ocess’. Capi al & Class, 1(1):
43–57. h ps://doi-o g/10.1177/030981687700100104.
31
Gabo ieau, Da id. 2012. ‘« Le nez dans le mic o ». Répe cussions du a ail sous commande ocale dans les en epô s
de la g ande dis ibu ion alimen ai e’. La nou elle e ue du a ail, no. 1 (Decembe ). h ps://doi.o g/10.4000/n .240.
32
Massimo, F ancesco S. 2020a. ‘Bu oc azie Algo i miche. Limi i e As uzie Della Razionalizzazione Digi ale in Due
S abilimen i Amazon’. E nog a ia e Rice ca Quali a i a, no. 1/2020: 53–78. h ps://doi.o g/10.3240/96824.
18
The p o esso s essed he poin ha an eme gence o a dynamic Eu ope o add ess ‘ us wo hy
AI’ was g owing, wi h inc easing esea ch. We need o c ea e undamen al and applied esea ch
posi ions wi hin he indus y o edisco e he link wi h he gene al public because we see gaps
being made, she men ioned.
Simila ly, ollowing an in e iew wi h a F ench lawye who co- ounded a i m specialised in AI
echnologies, he explained ha he legal wo ld is slowly in eg a ing algo i hmic managemen in o
hei ac i i ies o asks ha can be au oma ed. These include adminis a i e asks, con ac s,
o ecas ing and clien ela ions.
Un il ecen ly we had an old way o doing ad ocacy and now we a e s a ing o
make new mo es o make i mo e echnological, and mo e mode n. This allows
small law i ms like my own o limi low alue-added in e ac ions and use machines
o inc ease e iciency ins ead.
Law i m specialized in AI echnologies, Lawye
Al hough he i m’s main cus ome cases a e on da a p i acy and blockchain, hese opics we e
ound o be ele an o he logis ics indus y. Indeed, he e a e many legal p oblems wi h AI ools
and algo i hmic managemen , such as ques ions on da a use and p i acy o ques ions su ounding
esponsibili y and liabili y. Fo example, he lawye desc ibed he dilemma o esponsibili y e e ing
o he human behind a speci ic algo i hmic managemen ool, he manu ac u e o he p oduc , and
he designe o he algo i hm. The lawye said he e has been a g owing conce n abou he liabili y
o algo i hmic managemen ools ac oss he logis ics sec o . He said ha i a echnological p oblem
a ises om he use o an AMT ha has been ou sou ced o ano he company, he esponsibili y will
no always be clea . The e a e impo an legal and e hical issues, whe he he o iginal company o
he ou sou cing company is accoun able.
7 Su eillance and da a p i acy issues in he e a o algo i hmic
managemen
AMT may ack wo ke s' mo emen s and beha iou s, leading o inc eased su eillance, and, as a
consequence, inc eased p essu e, and s ess among he wo k o ces. Fo example, some companies
ha e epo ed ha he use o algo i hmic managemen ools such as handheld and wea able
de ices has led o inc eased su eillance o wo ke s, wi h GPS acking and eal- ime moni o ing o
pe o mance. This can c ea e s ess and anxie y o wo ke s and may lead o a mo e igid and
igh ly con olled wo k en i onmen . Simila ly, AI-powe ed ou e planning ools can also lead o
ex ensi e moni o ing and su eillance o d i e s.
Box 3. Case s udy on F ench Logis ics 4
A key applica ion o algo i hmic managemen ools in he logis ics sec o is concen a ed in T anspo
Managemen Sys em companies (TMS). F ench Logis ics 4 is a TMS so wa e company ha sells
So wa e as a Se ice (SAAS) h ough paying subsc ip ions a ailable in 4 languages and 6 coun ies,
wi h i s main ac i i y based in F ance. F ench Logis ics 4 c ea es he echnology behind a dashboa d
ha o he companies use o managing hei logis ics. Clien s include dis ibu o s o deli e y
se ices, ca ie s o cou ie se ices, o shippe s who manage hei lee using he solu ion, which
allows clien s o build hei logis ics pla o m. The company has de eloped hei own ou e
op imisa ion engine based on open sou ce.
Wi h he apid ise o deli e y se ices since 2016, ou i s clien s we e demanding a web po al o
easily manage hei logis ics, along wi h a s ong echnology expe ience ha many o he compe i o s
didn’ ha e. We hough why keep he echnology jus o us, bu ins ead make i a ailable o o he
deli e y playe s? This is how he ool was c ea ed, a new company, wi h mo e ech e o s o se e
people o in e egional deli e ies.
F ench Logis ics 4, Founde
19
The ool is pu ely a logis ics managemen so wa e ha allows use s o op imise hei deli e y ou es
acco ding o he d i e s loca ed nea by o o he chosen ac o s. F ench Logis ics 4 cus omises he
pla o m o he needs and hemes o i s cus ome s and ac s like an En e p ise Resou ce Planning
(ERP) o deli e y and logis ics companies, o o non-deli e y companies wishing o s a deli e y
se ices. The i s s ep in using he ool is o choose he deli e y imes, b oken down by day and hou ,
and delinea e he zone o ac i i y, which can be a neighbou hood, an en i e ci y, o a egion.
A e wa ds, he dashboa d gi es many op ions o use s o decide how o build hei se ices.
Cus ome s can choose hei desi ed se ices, a ea o in e es , he a i s pe km/minu e/dis ance, ype
o ehicle o be used, a ailabili y o se ices ( ime window ames) e c. Thei p ices can be se pe
km, pe base, pe olume o weigh . Figu e 3 illus a es he dashboa d o a demo company a e
ha ing chosen all o he abo e.
In his window, use s can ha e an o e iew o wha is going on pe deli e y, he missions
accomplished, hose done, dispa ched, ongoing, cancelled o ailed. Use s can also see he deli e y
missions, ime and loca ion o deli e ies, s a us, ime aken, numbe o goods deli e ed,
en i onmen al oo p in and many o he me ics. F ench Logis ics 4 does no do deli e ies bu only
sells he ool behind he dashboa d which allows companies o hen plan all hese opics. Each
company has i s dashboa d, and he deli e y wo ke s ha e hei applica ions connec ed o he
dashboa d.
Figu e 3: Example o dashboa d – lis o missions
Sou ce: Ma e ial p o ided by F ench Logis ics 4
Figu e 4 shows he ou e acking window in he ool’s dashboa d ha allows cus ome s o ack each
d i e , o all a once, and ecei e li e no i ica ions on hei s a us. This helps clien s manage hei
employees and inc ease in e nal communica ion om a cen alised a ea. Fo example, in an
eme gency scena io whe e a hospi al mus quickly ans e li e o gans om one poin o ano he ,
he ool allows one o isualise he s a us o he ip and i s emaining ime. This helps igge he
su gical s eps ea lie and no i y he doc o wi h ewe esou ces since all in o ma ion can be ecei ed
ia a phone app.
20
Figu e 4: Example o dashboa d – ou e acking
Sou ce: Ma e ial p o ided by F ench Logis ics 4
Figu e 5 shows he ool’s pe o mance o e iew window o all d i e s wi h a subsec ion o each
d i e . On op o he window, he cus ome can isualise he o al numbe o d i e s in he company,
he d i e s mobilised ha mon h, he a e age u no e pe d i e ha mon h, and he a e age
u no e pe d i e he pas mon h. A he cen e o he window, he cus ome can isualise he
e olu ion o new d i e s ac oss a ious mon hs and he op 10 d i e s wi h he mos missions
comple ed. A he bo om o he window, he cus ome can see he lis o d i e s wi h each s a is ic
as no ed abo e.
Figu e 5: Example o dashboa d – pe o mance o e iew o d i e s
Sou ce: Ma e ial p o ided by F ench Logis ics 4
21
The impac s o using he ool a e huge o cus ome s, said he ounde . Fo some e aile s, i has
allowed hem o en e he ma ke and s a deli e ing. Since Co id, many e aile s ha e ealised he
impo ance o o e ing deli e y se ices o hei business model, and Co id has boos ed cus ome
demand o deli e y. Fo o he s, i allows hem o launch hei business as e . The ounde no ed
ha hei cus ome s can go h ough hem o quickly s a hei deli e y business, which speeds up
deli e y en ep eneu ship.
Fu he mo e, o o he companies, i has enabled a dema e ialisa ion p ocess because, be o e he
in oduc ion o ools like his, deli e y was based on a pape wi h in o ma ion needing a s amp o
e u n o he o ice o s o ing and scanning. Finally, o la ge logis ics companies p esen in se e al
coun ies, he so wa e has highe s akes o hose wan ing o ha e he capaci y o manage all o i s
lows in a single ool. Mos companies wan o be able o check he p og ess o ope a ions in a ew
clicks.
The main impac o using he ool is sa ing ime on deli e ies and he e o e imp o ing he
o ganisa ion o wo k and people. The ool sa es ime because cus ome s no longe ha e o call he
deli e ymen o ecei e upda es as hey can easily loca e hem on a map. Less ime is was ed on he
elephone so unp oduc i e oles a e elimina ed, which is a inancial gain ha cus ome s can use
elsewhe e. In e ms o wo k o ganisa ion, i allows cus ome s o gain p oduc i i y because i equi es
ewe people o do mo e hings. Likewise, i is supposed o enhance easie communica ion be ween
he wo ke s and hei supe iso s.
When asked abou possible issues su ounding d i e s’ au onomy, as hey a e cons an ly being
moni o ed h ough he GPS applica ion, he ounde said ha esponsibili y was o each cus ome .
No mally, deli e y d i e s ha e minimum eedom, he said, bu a e ha , we canno necessa ily
know wha he conc e e applica ion is. When asked abou ela ions be ween business ac o s and he
in e ac ions be ween playe s in he sec o , he ounde said he ollowing.
In he luidi y o ela ions be ween playe s in he sec o , as he dema e ialisa ion o he sec o is slow,
dis ibu o s a e somewha a he me cy o he anspo e s’ ools. I I am a company and I ha e
ce ain dis ibu o s wi h di e en le els o digi alisa ion, I canno o e he same cus ome expe ience
e e ywhe e I ope a e. Thus, we allow hem o p o ide hei ca ie o use ou echnology whe e hey
wo k. This allows us o wo k mo e uni o mly. We connec dis ibu o s and ca ie s who use ou
so wa e.
F ench Logis ics 4, Founde
Rega ding he u u e o AMTs in he sec o , he ounde said ha hese ools will become inc easingly
essen ial o keep up wi h he pace o deli e y i companies wan o emain compe i i e. The e is a lo
o so wa e on he ma ke , some o which is old, he said. The TMS sec o is becoming e y
concen a ed wi h he akeo e o s a -ups and he use o echnologies depends on he way each
one ope a es.
Mo eo e , o ga he da a on he egula o y challenges o algo i hmic managemen ools, an
in e iew was conduc ed wi h he F ench Na ional Commission on In o ma ics and Libe y (CNIL).
Th ee expe s we e in e iewed, including a wo ke a he CNIL’s digi al labo a o y (LINC), a wo ke
in he solida i y and employmen se ice, which is pa o he legal suppo depa men , and a
sociologis a LINC.
The expe wo king in he legal depa men said AI ools a e pa icula ly expanding in HR and he
hi ing p ocesses o all companies. Algo i hms a e inc easingly used o applica ion so ing o assis
ec ui e s in analysing audio o ideo s eams and unde s anding whe he candida es a e ne ous.
The subjec is e y popula and o in e es o s akeholde s, e en hose wi hin Logis ics. CNIL has
published a guide on ec ui men applica ions wi h ce ain shee s on he use o algo i hmic
managemen . (h ps://www.cnil. / /le-guide-du- ec u emen : shee 4 on consen and shee 13 on
algo i hmic p ocessing ools).
22
The no ion o algo i hmic managemen is o au oma ed decision-making. In
Amazon, we see ha he p inciples o he gig economy becoming mo e and mo e
used o logis ics. We can see how he ‘ube iza ion’ o wo k applies o mo e and
mo e sec o s, including logis ics, whe e his o ically his was al eady p esen . O en,
we a e asked abou su eillance issues a wo k, which has become possible wi h
widesp ead ideo su eillance ools. These a e no only used o moni o a place bu
also people, and since su eillance ools a e inexpensi e, e en SMEs use hem wi h
a lack o knowledge o he legal amewo k.
CNIL, AI egula o y expe
The issue o wo ke su eillance emains a cen al conce n, wi h se e al high-p o ile cases d awing
public a en ion. Fo ins ance, he CNIL ined SAF Logis ics €200,000 o excessi e da a collec ion
om employees, iola ing hei p i acy and ailing o coope a e adequa ely wi h CNIL se ices
33
.
Simila ly, Amazon F ance Logis ique ecei ed a €32 million ine o implemen ing an o e ly
in usi e employee moni o ing sys em, as well as o using ideo su eillance wi hou p ope
no i ica ion o su icien secu i y measu es
34
. Addi ionally, du ing he Co id-19 c isis, he CNIL
played a key ole in con o e sies be ween employe s and unions ega ding wokplace su eillance.
One o he mos signi ican cases in ol es Amazon F ance (see Box below).
Box 4. A con o e sy on an AI-based sys em o social dis ancing a Amazon F ance
Logis ics du ing he Co id-19 pandemic (2020)
Du ing he Co id pandemic, Amazon de eloped new AI-assis ed moni o ing ools. One o hese was
P oxemic a so wa e o isual ecogni ion ins alled on came as, whose ask is o enhance social
dis ancing by signalling he c owding o wo ke s in he wo kspace. This was an a i icial in elligence
sys em ha analysed images om special secu i y came as and ale ed managemen o po en ial
social dis ancing iola ions. P oxemics was buil by AI expe s in Amazon obo ics di ision and
deployed in mid-Ma ch in he US. I was in oduced p og essi ely in o he Eu opean coun ies (1000
Amazon buildings a ound he wo ld acco ding o Wi ed). The sys em consis ed o a ele ision sc een,
dep h senso s and an IA-enabled came a, which a e ins alled in di e en poin s o he plan s. The
came a egis e ed images in eal- ime, acking people mo ing h ough he wa ehouse. When hey
pass in he isual ield o he came a wo ke s appea on he sc een su ounded by “augmen ed
eali y” ci cles. The IA used he appa en size o people in he ame and he numbe o pixels be ween
hem o calcula e dis ance. I he social dis ance is espec ed, ci cles a e g een; o he wise, ci cles a e
ed, a possible iola ion is lagged, and managemen is ale ed. Re iewe s include he de ails in a
egula epo sen o building manage s ha summa izes ecen social dis ancing iola ions in hei
acili y. These ea u es o P oxemics aim, Amazon says, o p o ide a quick esponse o con agion isk
and a e used only o Co id-19 sa e y. Howe e , his sys em aised many conce ns abou p i acy on
he pa o unions and o he independen obse e s, as well as egula ion au ho i ies ( o ins ance in
F ance).
33
CNIL. 2023. Excessi e da a collec ion and lack o coope a ion: he CNIL imposed a sanc ion on he company SAF
LOGISTICS. h ps://www.cnil. /en/excessi e-da a-collec ion-and-lack-coope a ion-cnil-imposed-sanc ion-company-sa -
logis ics.
34
CNIL. 2024. Employee moni o ing: CNIL ined AMAZON FRANCE LOGISTIQUE €32 million.
h ps://www.cnil. /en/employee-moni o ing-cnil- ined-amazon- ance-logis ique-eu32-million.
29
Lis o boxes
Box 1. Case s udy on F ench Logis ics 1 ............................................................................................................................... 11
Box 2. Case s udy on F ench Logis ics 3 ............................................................................................................................... 13
Box 3. Case s udy on F ench Logis ics 4 ............................................................................................................................... 18
Box 4. A con o e sy on an AI-based sys em o social dis ancing a Amazon F ance Logis ics du ing
he Co id-19 pandemic (2020) ......................................................................................................................................... 22
30
Appendix
1. Dis ibu ion o in e iews
Table 1: Dis ibu ion o in e iews.
In e iewee
Role/O ganisa ion
Da e
HR and Social
Rela ions Di ec o
Anonymous - mul ina ional
shipping & ecei ing and supply
chain managemen company
01/12/2022
Accoun and P ojec
Manage
Anonymous - so wa e and
consul ing i m
01/12/2022
9 di e en
echnology expe s
and p ojec leade s
Anonymous –Independen
logis ics and supply chain
associa ion
02/12/2022
Co- ounde s and
Pa ne s
F ench Logis ics 1
09/02/2023
Lawye , P o esso
and Tech expe
F ench law i m
09/03/2023
P ojec o ice
F ench logis ics ounda ion
10/03/2023
‘AI Secu i y’ g oup
leade
F ench logis ics associa ion
13/03/2023
CEO
F ench Logis ics 3
13/03/2023
Di ec o and Tech
Expe
F ench Logis ics 2
16/03/2023
Da a Scien is
Anonymous - mul ina ional
con aine anspo a ion and
shipping company
17/03/2023
Compu e Science
and AI E hics
P o esso
Pa is-So bonne Uni e si y
13/03/2023
Co- ounde
F ench Logis ics 4
24/03/2023
Technology and AI
expe s
Na ional Commission on
In o ma ics and Libe y (CNIL)
27/03/2023
Sou ce: Au ho s’ elabo a io
Ge ing in ouch wi h he EU
In pe son
All o e he Eu opean Union he e a e hund eds o Eu ope Di ec cen es. You can ind he
add ess o he cen e nea es you online (eu opean-union.eu opa.eu/con ac -eu/mee -us_en).
On he phone o in w i ing
Eu ope Di ec is a se ice ha answe s you ques ions abou he Eu opean Union. You can
con ac his se ice:
— by eephone: 00 800 6 7 8 9 10 11 (ce ain ope a o s may cha ge o hese calls),
— a he ollowing s anda d numbe : +32 22999696,
— ia he ollowing o m: eu opean-union.eu opa.eu/con ac -eu/w i e-us_en.
Finding in o ma ion abou he EU
Online
In o ma ion abou he Eu opean Union in all he o icial languages o he EU is a ailable on he
Eu opa websi e (eu opean-union.eu opa.eu).
EU publica ions
You can iew o o de EU publica ions a op.eu opa.eu/en/publica ions. Mul iple copies o ee
publica ions can be ob ained by con ac ing Eu ope Di ec o you local documen a ion cen e
(eu opean-union.eu opa.eu/con ac -eu/mee -us_en).
EU law and ela ed documen s
Fo access o legal in o ma ion om he EU, including all EU law since 1951 in all he o icial
language e sions, go o EUR-Lex (eu -lex.eu opa.eu).
EU open da a
The po al da a.eu opa.eu p o ides access o open da ase s om he EU ins i u ions, bodies
and agencies. These can be downloaded and eused o ee, o bo h comme cial and non-
comme cial pu poses. The po al also p o ides access o a weal h o da ase s om Eu opean
coun ies.
The Join Resea ch Cen e (JRC) p o ides
independen , e idence-based knowledge and
science, suppo ing EU policies o posi i ely
impac socie y
EU Science Hub
Join - esea ch-cen e.ec.eu opa.eu
