Armin Beverungen, Ulf Treger, & Maja-Lee Voigt (Leuphana University Lüneburg) presenting “Justice at the End of the Supply Chain: Interrogating Amazon’s Logistical Urbanism from the Cloud to the Curb” (2/21/2024)
—
Resources
* Abdelhadi Belfadel, Sebastian Hörl, Rodrigo Javier Tapia, Dimitra Politaki, Ibad Kureshi, et al.. A Conceptual Digital Twin Framework for City Logistics. Computers, Environment and Urban Systems, 2023, 103, pp.101989.
* Altenried, Moritz. “On the Last Mile: Logistical Urbanism and the Transformation of Labour.” Work Organisation, Labour & Globalisation 13, no. 1 (2019): 114–29. https://doi.org/10.13169/workorgalaboglob.13.1.0114.
* Alimahomed-Wilson, Jake. “The Amazonification of Logistics: E-Commerce, Labor, and Exploitation in the Last Mile.” In The Cost of Free Shipping: Amazon in the Global Economy, edited by Jake Alimahomed-Wilson and Ellen Reese, 69–84. Pluto Press, 2021. https://library.oapen.org/handle/20.500.12657/46326.
* Anderson, Chris. The Long Tail: Why the Future of Business Is Selling Less of More. 1st ed. Professional Development Collection. New York: Hyperion, 2006.
* Berlin VS Amazon (2024): Material. https://berlinvsamazon.noblogs.org/material/ [last accessed 21.02.2024].
* Cook, William: In Pursuit of the Traveling Salesman: Mathematics at the Limits of Computation. First paperback printing. Prinston, Oxford: Princeton University Press, 2014.
* Dayen, David (2023): The Lie Behind Amazon’s HQ2 Sweepstakes Becomes Clear. The American Prospect. https://prospect.org/api/content/b6fe284c-ba13-11ed-90b9-12b3f1b64877/ [last accessed 21.03.2023].
* Delfanti, Alessandro. “Machinic Dispossession and Augmented Despotism: Digital Work in an Amazon Warehouse.” New Media & Society 23, no. 1 (January 1, 2021): 39–55. https://doi.org/10.1177/1461444819891613.
* Von Wegen Lisbeth (2019): – Alexa gib mir mein Geld zurück! (Offizielles Video). https://www.youtube.com/watch?v=1PJvR0qjZIE (last accessed 21.02.2024)
Illustrations by Maja-Lee Voigt. If not marked otherwise, the pictures were taken by the Automating the Logistical City Research Team.
—
Armin Beverungen is junior professor for organization in digital cultures at Leuphana University Lüneburg, where he is also a member of the Centre for Digital Cultures. He is an associate editor of Organization. His research currently focuses on two projects: on how Amazon reshapes our cities (https://logistical.city/, with Maja-Lee Voigt and Ilia Antenucci), with a particular focus on automation, logistics and speculation in the city; and on the promises of wealth associated with smart urbanism (with Orit Halpern, Marc Steinberg, Liza Cirolia and Anindita Nag).
Ulf Treger (he/him) is fellow with the Center for Digital Cultures and associated with the project “Automating the Logistical City”. He has studied Visual Communication at HFBK Hamburg. As a designer, coder, and researcher he works on projects on the interdependence of digital and physical spheres and researches their impacts on social practices and public space. One focus are maps as visual representations of space and also as instruments of creating spaces.
Maja-Lee Voigt (she/her) is an urban researcher, Ph.D. student at the Leuphana University Lüneburg, and co-founder of the interdisciplinary city research collective Akteurinnen für urbanen Ungehorsam in Hamburg, Germany. Assisted by a methodological toolbox of ethnographic and critical feminist thinking, she is currently researching Amazon’s monopoly on bits, bytes, and boxes. Overall, Maja-Lee’s work focuses on the automation of logistical cities, tackling questions about (resisting) algorithmic architectures of oppression, and hacking patriarchy towards more just urban futures.
DELIVERY EXCEPTION: SUPPLY CHAIN JUSTICE & RECONCILIATION is a speaker series bringing together scholars and organizers to discuss logistical justice and examine the possibilities of reconciliation in an era of supply chain capitalism.
All right uh we’ll go ahead and get started hi everyone welcome to delivery exception uh this is a short speaker series we’ve organized for this spring uh to bring together Scholars and organizers to discuss logistical Justice and the possibilities of reconciliation in an era of supply chain
Capitalism uh this event which is free and open to the public is uh also organized as part of the supply studies research network uh a group of Scholars students activists and organizers uh who have identified an active interest in the critical study of logistics uh with the research Network
Itself supported in part uh by the National Endowment for the Humanities office of digital Humanities funding an additional support for this series is provided by the conference of Arts and Science Deans and the department of communication and media studies here at forom and while as a virtual event the
Geographies that are being enrolled in our uh assembly here today are many I’d like to acknowledge that forom itself occupies the ancestral lands of the Lenny lenapi and wer peoples uh before I introduce our speakers for today i’ just like to go ahead and uh do the schedule for the
Rest of our events this spring so our next event is Wednesday March 6 at 7 pm uh with Christina denbar Hester from the University of Southern California an Athena tan from plugin IE that IE is Inland Empire uh Monday March 18th at 6:00 p.m. we have Miriam Posner uh from
The University of California Los Angeles and then Wednesday March 27th at 6m we have Benjamin mcken from Ohio State University Jessica champagne from the worker rights Consortium and an Hela solise from make the road action so I hope to see you at some of these future
Events so uh like to go ahead and introduce our speakers the title of their presentation today is Justice at the end of the supply chain interrogating Amazon’s logistical urbanism from the cloud to the curb our speakers are uh Armen bangan uh ol Trager and Miley Vogt uh Armen is a
Junior Professor for organization in digital cultures at lefun University Lindberg where he is also member of the center for digital cultures he’s an associate editor of organization and his research currently focuses on two projects on how Amazon reshapes our cities uh and that’s with a particular focus in automation Logistics
And speculation in the city and on the Promises of wealth associated with smart urbanism Al Trager is fellow with the center for digital cultures and associated with the project automating the logistical City he studied visual communication at hfbk hurg as a designer coder and researcher he works on projects on the interdependence of
Digital and physical spheres and researches their impact on social practices in public space one Focus are maps as visual representations of space and also as instruments of creating spaces Miley V is an urban researcher PhD student at the Lana University of Linberg and co-founder of the interdisciplinary City research
Collective um I’m going to do the translation into English which is I think actors for uh Urban disobedience in uh hurg Germany uh assisted by a methodological toolbox of ethnographic and critical feminist thinking she is currently researching Amazon’s Monopoly on bits bites and boxes overall my’s work focuses on the automation of logistical
Cities tackling questions about resisting algorithmic architectures of Oppression and hacking patriarchy towards more just Urban Futures so welcome to you all thanks so much for being here and I’ll go ahead and hand it off to you uh attendees we have time for questions afterwards uh you can either
Uh paste your question as we go in the Q&A button at the bottom of zoom and we’ll answer it afterwards or um after the speakers are done uh you can go ahead and raise your hands and I will call on you and I can activate your microphone to allow you if you want
To uh you know uh speak your question you’re welcome to also type it into chat or in that Q&A panel so uh without further Ado please go ahead yeah hello and welcome also from me from us it’s a pleasure to U be part of this uh wonderful um set of speakers
And part of the series thanks for having us Matthew um and it’s fun to be part of the the network uh and um yeah and it’s been um actually quite a joy to put together this presentation that will should take us the next 45 50 minutes or
So um yeah we’ve already been introduced so I’ll I’ll start off with giving you a bit of a background to where this all comes from and the background is really a research project on automating the logistical City this is um the the front page of our website and we are concerned
In that project um with studying how Logistics and its algorithmic algorithms transform Urban spaces and turn them into sites of ordinary specul so particular focus of the project is on Amazon as you would have spotted already which we think has been granted too little attention is a force really
Remaking the urban beyond the impact of its search for new headquarters the building of signature skylines or more broadly the decline of retail and the rise of delivery and we really spent the last two years attempting different inroads into studying Amazon and its impact on the urban focused in
Particular on Hamburg and on the ru is that okay presentation um which you can see on the next slide just so you know where we are um in Germany um so this is a bit of a different story I guess a bit of a different narrative perhaps from what
You’re used to if you’re based in the US or in other places um and we’ve understood this um work as an attempt to open up the black box that Amazon largely still is starting from the premise that our methods were largely involved workarounds as Amazon is not likely to welcome our research efforts
With open doors we’ve had I think three interviews with Amazon reps and that’s about it um so this kind of turned out to be true and instead we’ve deployed a multiplicity of methods to come to grips with how Amazon shapes this city um our research has focused mainly
On two empirical sites as I mentioned um so Hamburg is a city of Northern Germany dominated by by its large Port of media industry and um the rural area is um a post industrial area which is embraced Logistics after the decline of the coal and steel industry and we’ve
Complemented this Regional focus with broader Research into Amazon’s platform infrastructures for delivery also in its relation to Amazon web services as the tax stack if you like for much of its operations and our research included uh things like fulfillment center and Port tours visits of Amazon sites um from
Fulfillment centers to delivery stations to Amazon lockers and attempts to map how these infrastructures of logistics have impacted the city it is also included in in a kind of a science and technology studies register uh studying patents and exploring Amazon science blogs and activities such as its last
Mile routing research challenge that all will talk about and of course there’s been loads of interviews with some Amazon reps um labor union organizers and other experts working in and around Urban Logistics and overall this multi method research has produced a very patchy um still insight into Amazon
Urbanism which we would like to present today in three parts and I will start mostly talking about Labor um questions of Justice in relation to labor ol will continue with talking about algorithms and algorithmic J Justice and my will conclude talking about space and spatial Justice and of course these are highly
Related and we’ll point out to some of point out some of those connections and we will each start with a little vignette from the field to give you a bit of a feel of our research um also um needs to be said that Ilia antonucci our colleague on the project unfortunately
Can’t be with us today but if you’re watching this later hello IIA uh okay so I’ll start now with a um a quick field note from Amazon I’ll just read this out um so this is from a field trip to a location in Dortmund um last September
That I um to did with our colleague Andrea polio who’s based in Turin uh and I think you can put the next slide already um we’ve taken the route through the main entrance of the old V Fen hitter situated Northeast of Dortmund City Center passing underneath a Railway
Bridge still emblazoned with t tsen cop steel and large LS the site used to be one of the largest steel plants in the world until its decline in the 1990s in the dismantling of the centering plant and other parts and which after 2002 were sold and shipped to um shanang in
China to be rebuilt there now only one steel coting plant and one roning B left and otherwise the vast site has been redeveloped to mostly host logistical Enterprises these include two Amazon F filment centers one of them Amazon’s first um what they call inbound Cross Dock in Europe and through which used to
Pass more than half of all goods sold by Amazon in Europe until more of the same type of fulfillment center were established elsewhere but we’re not here for the Fulfillment center instead we’re here to investigate reports of a new Amazon delivery station which was meant
To have opened already a year ago at the end of 2022 we approach the two new vast Logistics Halls which stand in an 80 Dee angle to each other built and owned by the logistics property developer prologus you can see them kind of um between the two railway lines in the top
Right there those are the the two top top ones there if you can make out where they are the business sign next to the entrance lists one tenant for the second Hall Amazon DN Z3 as we continue up the driveway we’re greeted by dozens and dozens of Teslas however some of which
Have QR codes attached to the windows which with a prompt you can buy me apparently the other Hall has been turned into an ad hoc Tesla delivery Hub a small Hut at the entrance gates to the parking wall coming potential buyers as we turn right on the parking lot we
Notice at the near end of the second Hall seems completely empty all loading bites are shut and closed and there are no logos on the side of the hall except for prus as we drive further down the vast lot we encounter a sign that announces Amazon DNZ 3 yet the parking
Lot is earily empty until we notice that the last loading Bay of the hall is half open with a Volkswagen Golf parked in front of it next to it at the very end of the hall there’s a recreation area where it seems someone is taking a break
We leave the car check out the bike stand and Survey the area there directly next to the parking lot newly created bike lane leads back to the notat the northern part of Dogman where traditionally workers at the steel plants used to live and today still migrants arrive and find their footing
In this Multicultural City the parking lot as well as the loading area in front of the gate seem hardly used there are no tire marks on the concrete except presumably from a few Joy Riders circling around the screeching tires at night we approached the six men most of
Whom are playing table tennis one of them sat on the back rest of a bench smoking we’re told that um there the engineering team dismantling the equipment from the delivery station apparently the station was all set up for operations a year ago but that was never put to use the team is now
Dismantling and packing up the equipment to be shipped somewhere to somewhere near Breman perhaps oldenborg in the north of Germany where we later found out a new delivery stations is to open soon go to the next slide so that was the the field note um
So this field note from our visit to the site and dment documents in some ways how speculative Amazon still velopment in its last mile delivery network is it quickly opened a lot of delivery stations and then noticed it had excess capacity leading to not even opening
Some of the stations it also alludes to the kinds of spaces Amazon aners as it develops its logistical infrastructure including the case in the case of dotman and many others a post-industrial site in close proximity to neighborhood with loads of Migrant labor Amazon’s fulfillment centers are well established
In Germany which is Amazon’s second largest market worldwide and what is more recent those it’s venturing into last smile delivery uh with heavy expansion taking place since 2017 and it has since started operating more than 70 delivery stations where Parcels are sorted for delivery to homes in the beginning it relied quite heavily
On Amazon Flex so on selfemployed labor where the key skill required is a driving license and which can be scaled up very quickly and complement more uh firm Labor Relations um to kind of fill up holes and especially for kind of late light delivery however Amazon shut down
Amazon Flex in Germany in the summer of 2022 and now relies pretty much solely on subcontractors so-called Delivery Service Partners to complete their last smile on the right here you can see um a graphic by by Maya Lee which shows you how how this works you go from fulfillment centers to sometimes salting
Cent not always and then the trucks arrive at the distribution centers where um things are you know sorted for final delivery before the Vans hit the road and bring us our pass us through a smile uh next slide please my um to understand labor at Amazon which is what
I’m supposed to be talking about um we need to acknowledge the different kinds of labor and employment relations that Amazon employs so this would be perhaps familiar to you but I thought I’d run through it anyway um Amazon deploys these different ways of including labor and its logistical operations um the
Situation in fulfillment centers I think is well documented so for setup Amazon often relies on temporary work agencies in Germany and elsewhere who provide kind of the first batch of of of Labor there’s usually then a transition towards uh contracts with Amazon often starting a seasonal labor in advances in
Advance of the Christmas uh period and this comes with the promise of being offered a permanent contract uh which is of course often broken as people are made redundant on the day before Christmas but those who are allowed to stay on remain on the maximum possible probationary period of two years before
They finally get a safe permanent contract and for last mild delivery though Amazon relied somewhat on Amazon Flex um is as I said a scalable self-employed labor but now relies pretty much completely on subcontracted labor employed by these Delivery Service Partners and this effectively means that Amazon can delegate responsibility for
Labor relations in The Last Mile while effectively imposing its contracts so Delivery Service Partners apply to Amazon must conform to Amazon’s conditions for example in relation to the availability of backup drivers or making themselves available for you know contact even sitting in these stations during the day is they’re monitoring the
Drivers that’s the only way they can be actually part of this delivery service partner program especially what they call the DSP 2.0 model involves Partners working exclusively for Amazon so they’re really very dependent on them um there’s there’s so much to say about labor at Amazon Labor Relations I just
Wanted to pick up two things that became very obvious in um in working through looking at the Fulfillment centers but then also the delivery stations and the last M the first first one is um algorithmic integration which is the next slide um and what’s um what’s kind of interesting
Here is that the kinds of algorithmic management taking place in fulfillment centers which are well documented are also kind of spread Beyond so you may um know Alexandra Dean’s work um for example who writing about um the Fulfillment sentences noticed that we have this something he calls machinic
Dispossession which is um when you have this chaotic storage where stuff is just put all over the Fulfillment center nobody knows where stuff is only the machine knows and this is kind of this um he calls it Machining dispossession to note how um you know it it um withdraws the ability to actually
Control work environment and we also have documentation of all these forms of surveillance um so not so much visual surveillance as data veillance and capture which Delan calls a kind of augmented despotism because it really leaves you um subject to all sorts of um managerial prerogative and well random prerogative
What’s Starling about delivery labor is how it is equally tightly integrated so in a fulfillment center it’s obvious maybe because you’re you know you’re in this building that Amazon made to uh control your labor um but it also works uh in pretty much the same way when you
Uh when you write when you deliver for Amazon so the flax app which is the you know which is the ones that self-employed drivers use um is also actually one that all the other drivers for the Delivery Service Partners use so the apps both provides information on
Shifts are the blocks that um the drivers are assigned to and it also provides routing and navigation uh information and of course all of this is very tightly um used to monitor performance so um there’s little scope here for decision- making by drivers in the same way it is in the Fulfillment
Center I’d say and the routing is calculated centrally so this actually happens uh in for all of Europe even in one kind of operation Center in in Barcelona is one of the interviews that myy interviewed told us uh and drivers must adhere to the to the routes and to
The order of delivery which is provided in that way because you know this is what machine learning wants otherwise the machines can’t learn what you see here on the slide is a dashboard for Delivery Service Partners and this is how both Amazon and the delivery um Service Partners track drivers so this
Is from a from a video on YouTube we didn’t get an insight into the software but um it’s it’s fairly obvious that you know this is how um kind of GPS tracking Works where um all of the drivers um you know when they use their apps um are
Trackable very quickly and and it means that the uh The Delivery Service Partners um next slide please so that’s the one one thing that’s maybe um worth mentioning this kind of algorithmic management that spreads from the Fulfillment center to the delivery station and to um to the road the other
Thing is abstract localism is I I don’t know we haven’t really found a better term for it yet we haven’t thought about a better one and this is this weird way in which Amazon clearly has this uh kind of very standardized managerial techniques that it that it uses especially the the the algorithmic
Techniques of um uh you know of of routing um of of of surveillance and so on but also and and then also these kind of cultural forms of management like all of the Amazon um principles that you see in blazing kind of all over the delivery stations and fulfillment centers but it
Also does this kind of um uh dealing with kind of the local meaning of Labor so this just two examples here one of them on the right is from a a sorting station in in in vitten adment and you can see what you see it says DTM 9 the
Sort Center and then what you see in the middle is the the classic pick of the coal miner and the classic helmet of the coal miner with a lamp in front and of course this alludes to kind of very very established traditions of kind of um um
Labor where you know where labor is highly valued makes you proud to work you know the co Miner used to be very proud and that this is kind of meant to be transferred to working for Amazon so it kind of picks up in a in a very crude
Way however you know every delivery station gets one of these signs um tries to pick up on these on these histories and cultures of of Labor in order to kind of um value labor and and motivate staff and then on the other side you see this um this is from inside the delivery
Station in vatal um south of kind of 50 km south of Dortmund uh on diversity um whenever you walk into any of these places and watch any of these videos it’s always about how diverse these places are and of course that’s that has to do with uh kind of the migrant labor
Force that’s um that’s uh that actually works at these places um that have to be kind of motivated to actually embrace uh the workplace which is part of its diversity so so this is kind of a irr response I think to local conditions both in terms of the valuing of Labor
And the the migrant labor force the divers labor force okay next slide and when we think about Justice then um because this is in the title of the uh of of of the series is and there’s there’s there’s so much to say I just want to um point out three things
There’s obviously a huge in collective bargaining there’s a huge debate in Germany and this is interesting compared to the US that it’s all about um whether the bargaining um the um The Benchmark is retail or Logistics because in Germany the logistics um Collective bargain Agreements are a lot
Worse than the ones for retail um but Amazon says it’s a logistics company here rather than a retail company so it pays Logistics uh well roughly Logistics wages not retail wages so this has been an ongoing one of the major ongoing demand from the unions to for Amazon
Tecknowledge as a retailer and that needs to pay retail wages um there’s also all of these kind of Outsourcing subcontracting self-employment forms of Labor Relations which we can call kind of the Amazon ification of um uh of of logistics um which uh rais loads of questions about you know basically
Amazon evading all sorts of responsibility for Labor Relations there’s also huge problem or challenge to kind of organize Last Mile because the unions really aren’t getting to the last mile they get to fulfillment center some delivery stations although they’re too new um but they’re not really
Getting to the the large um labor force migrant labor force and delivery which is also has a high turnover and is outsourced so that’s kind of a a huge uh well stumbling block when we Tred more to think about Justice and then on the right you see a poster from a art group
Called nootica and algorithmic Justice which means I’ll now pass on to olol to continue thank you I will also starts with a f note and I will read it from I think a f notee of Maya Lee it’s hws Meetup organized by Moya July 6 2023 Hamburg it’s pretty crowded around 40 to
50 people around here my neighbors unfortunately are just here to meet people and learn more about hws Amazon web services so no expert yet there’s free pizza and drinks dudes with funny t-shirts some hipsters classes white sneakers one pair of different stocks mostly casual clothing the event is
Generally in English and one out of three female representing persons excluding organizers no one is taking notes barely any pictures are taken of the slides the Moya location is very stereotypical tech office with the common area including a big screen everyone is assembled around it a big open kitchen area with Ikea looking
Lamps out of straws colorful walls everything seems very pride and fun big lettering on the bathroom doors age of people mostly in the 20s and 30s The Talk itself is not super beginners friendly I’d imagine lots of specific Rock up but maybe better understandable with the developer background I wonder
If the speakers get any hws discounts for the talk I’m thinking no there are two questions during the talk asking about specifics some Snicker sometimes add the glitches my neighbor has Amazon app on his phone he sells t-shirts and pants and has an online shop he looks even
Even younger than 20 maybe he uses the by Amazon yeah my task uh the second part of our talk um will be to talk a little bit about our efforts or recent efforts to scratch a little bit on the Black Box I mean I mentioned it before that um
Like pretty much all large big Tech platforms Amazon is a black box where it’s very difficult to take a look side especially if you take a talk about aloric um control and um steering of all the Enterprises next slide please um One um main theme and the
Research is um mentioned before um to research a little bit more about the last mile which mean which means from the lightest delivery or sorting Center to the doorstep of the customers and our effort was to find out a little bit more how the algorithmic work is controlled
And um um um organized around one question which is here typically cited by an quote of Chris Anderson Chris Anderson wrote in 2006 this um fames book um the last not the last M but the long tail which kind of like strategic background uh description of what Amazon drives up today and
There’s one small chapter about logistics and this chapter has the title of the tyranny of geography and it somehow somehow describes very well the Viewpoint of Amazon of the city of The Last Mile that this is like something that has to be overthrown or competed or conquered and for get the control next
Slide please so the basic problems on the perspective of Amazon are that the last mile is extremely time and Coast intensive it contains a lot of uncertainty the workforce the workers the drivers and the context of course the city or the urban context and therefore the last mil is
Probably the last controllable part of the supply chain maybe math can help here and we come back later to some things like traveling salesperson problem who could maybe optimize this problem and help to maximize profit and income next slide please as I mentioned it’s really
Difficult to get a view um in the P box we um recently tried three different approaches and I would like to roughly show a little bit what we’ve done the first example will be that we just wanted to recreate somehow the software that um Amazon is using for organizing
The whole um Last Mile um delivery and then um I will talk a little bit about the Amazon tsp challenge in 2021 and then a kind of like more playful speculative design The Last Mile um aloric system next slide please the um routing simulation the idea since we can’t look directly into
The black box is to look out how Amazon is perhaps reusing their own algorithms or the methods that are used in their own algorithms and how they perhaps make it accessible for others and there is Amazon web services we all um I think know that Amazon web services is a
Really huge cloud and service provider I think it’s getting much more money in the box of Amazon since the delivery itself and there there are some like light versions of the routing algorithms that are used by Amazon itself sold to the public monitor rise in extra way without having probably the most comfort
And accuracy they’re using for themselves but um we thought it would be worth to try out a little bit to use this public available Commercial Services to find out what um is possible with that so we made a really simple not really sophisticated application just to
Find out how the routing and the stuff and the data that comes from that work and how the conditions are that of that next slide please some preliminary insights are really basic so we found out that Amazon has a huge Universal catalog of places so every place um in in the areas where
Amazon is active has a hashed ID and this hashed ID is connected with sever um uh data sets of customers U public customers commercial customers and also um individuals there the routing and Maps which is the interesting stuff from the algorithmic or mathematical perspective are provided by external corporations
Here in um northern hemisphere it’s here Technologies or SRI in Southeast Asia it’s also crab Maps then there’s the route optimization so to calculate the optimal route so that no driver has to drive longer as it’s necessary it’s done by a tool which is open source from
Google and this is the assumption that leads us to the next questions there might be a lot of real- time updates to inform ahead before the right calculation happens about Ro blocks congestions and events that might disturb the delivery service so the question would be also where does this
Real time data comes from and next slide please here I just mentioned it is one of the two main map and routing information um provider for Amazon Amazon web services and here Technologies I don’t know if you know this Corporation is owned by five to six of the largest Automotive um
Corporations in the world for Germany for example is Volkswagen mercedesbenz and BMW they bought this Corporation from nuia some years before and um they’re running like navigational software for a lot of their cars and they’re also providing the services um for others like example for Amazon and here is um
Practically a little bit more open-minded to talk about their um business about their blackbox and they are doing like typical data exis so they’re get getting their real time data also the basic data for the place on world index but just scraping the worldwide web and to um collect
Every place of interest and information they can collect from the world weite next slide please the other main data source of here is public data and uh in the research we found a memoranda of understanding of Hamburg which is one of our two places of interest and um and
Where I am staying actually also and um here is a memorandum of understanding for 2017 where the city of hurg or the government of hurg agreed to share public available data and much more this here and hope in exchange to get some data from here next slide
Please so if you take a look a little bit of this public data from Hamburg just to mention the background Hamburg is like every other larger or smaller City struggling with digitalization um there’s this um construct of smart city or digital City Hamburg has a really big um digital um
Um City strategy since 10 years and they are also like investing a lot of doing and producing data also for strengthening the mobility um um sector and to provide um public data for new business models so to quote them and to startups in the mobility sector and to
Make it possible that a digital IM image of the city can be created and may be accessible for applications and potential users next slide please so just to give one possible use case Hamburg is offering um with their data in their Urban data Hub is this prototype which
Is called traffic lights data from hambur work it’s the realtime data of all traffic lights in the downtown area you can see this visualization you can see the crossings and the streets were leading to The Crossings and every street that are approaching to a um Crossing has the red light of the
Traffic light is here indicated in red and of course Green in the other direction if the road is free and this is just an example how uh um hambrook is introducing different kind of real time traffic data that could be used via their open data GE portal and that could
Be used also according to the memory of understanding by here and at the end of their logistical chain than by Amazon um um to optimize their routing and um their fulfillment next slide please this was one approach to find out um where the data comes from um during
Um by our simple um demo the other thing that was really interesting is an event that was initiated by Amazon 2021 it’s called the tsp challenge tsp stands for the traveling salesperson problem the traveling sales person problem is one of the most basic and most interesting mathematical problems which is discussed
And research since the 19 1940s and it’s basically the point to optimize routing and Logistics it comes from the idea that A salesperson is going to different cities to sale things and um that this person doesn’t want to get in one city more than once and of course it wants to
Have a route that is as efficient as possible the problem now is tsp it’s quite simple to explain it’s quite simple to um calculate if you like have two three or 10 stops but it’s getting really really complicated to calculate if you have for example 150 stops this
Is um like the measure this is mentioned in several interviews with Amazon workers or managers which is usable for um a worker delivery worker must bring out 150 things so the problem is exponential growing Solutions and it’s really difficult to find an exact solution next slide
Please so during the tsp uh challenge by 2021 by Amazon the winning team for example and many others who are working with the tribal sales person problem are not trying to get really the really perfect mathematically perfect solution they’re getting or they’re seeking to get an um Solution that’s quite optimal
It’s not perfect but in a reasonable time and um with not so much competing power um they are sufficient to get get a pretty good example so they’re using tics they’re using linear programming and other tools to get in a direction of that and this is somehow interesting
Because they are not in the pursuit to make it like really in a mathematical clean way but like in rough more um rock and roll way to go into this direction and to find a solution next slide please and this approach not to um be
Too accurate is a commod or um is coming together with um the idea that it’s really also really efficient to learn from real life examples so the winner team of the travel sales person challenge in 2022 um one um started to um really deeply investigate in U driving real driving scenarios and they
Try to find out patterns and transported or transformed the patterns into algorithms and that was the way they won the challenge so this like really an an interesting experience like as we learned before from I mean that algorithmic control of the delivery worker is somehow very um intense but
Also we all know that um the life in the streets in urban areas is um um um full of messiness of encounters and things that can happen so that um a normal dver still has some possibility to agitate um on her own on on his own to optimize his working
Routine and this re research goes in the direction to learn from this pattern and try to translate this into algorithms to somehow can mimic the real life experience of the ders next slide please so we have now the idea to bring all together this experiences not also
From the field research and um um the um excursions to the different centers and watching the uh Last Mile activities and also getting the algorithms somehow um explained a little bit more clearly we put together in a kind of speculative design all the components of the know
How we put together to get like a little bit clearer picture of that and just to summon up here really in the short time there are three main components of like any software and especially of this one these are the inputs so sensing for example I explained earlier the public
Data or using apis and other um formats to get data into the blackbox then there’s the processing optimization the traveling salesperson problem solving the time frames the Fleet Management the workforce organization and so on and then the output puts which means basically the rules that the drivers um receives via
The Amazon Flex app and um through other methods next SL please this is just small insight into the speculative design where we try to put together these different threats and um um experiences and assumptions we are doing to get a little bit a better picture of what’s going on next slide
Please and still there are a lot of questions open to like to um research more deeply um the question which which data source are used how local and temporal constraints are anticipated static or temporary ones what is done to prevent failure react to interference delays one hint is a traveling
Salesperson challenge um to learn from the drivers for the knowledge and how this can be translated into algorithms and how after all real life pattern can be transformed into rules and regulations next slide please um and this is of course a question a lot of researchers and Scholars are following there are quite
Some um papers has been released in the last two years by scholars in the direction of business studies um and they are like constructing also the picture of a more or less complete model where you can see here in a really um uh typical constellation in three columns
On the left hand side there’s the demand model which means um um the demand of the customers that should be fulfilled the blue one in the middle is the whole routing um problematic which I exlained earlier and the right side and yellow one is um the demand for real-time
Traffic data to bring the um uh delivery Goods as efficient as possible to the customers next slide please yeah and before we come to um um my Al part I just wanted to um like wrap up a little bit that I think in terms of um Justice
There are two points I think remarkable if you go into the direction of algorithmic control of the thing one thing is where the data comes from and in this case for example if data comes from free and public accessible um data sources the question is if this is um a
Way we should agree to or it would be better to regulate this somehow that will be perhaps a discussion for later and this the other point of Justice would be of course that um the workforce is exploited not only for the direct work but also for for tracking and
Learning from the behavior of those people to make smarter algorithms and perhaps in the future to replace thank you yes so we’ve heard a little bit of the front end of the logistics uh that Amazon is operating um and a little bit of the backend the algorithmic architectures that it those Logistics
Are based on um and the data and I kind of want to give you uh the bigger picture on what this does to our cities and what we call the logistical City I do want to start with a field note as well so uh bear with me um and give you
A little bit of an insight into one of the neighborhoods uh that Armen was already mentioning in Dortmund uh which is the NAD where a lot of workers come from or live um and where Amazon’s uh Distribution Center there is also based so this field note of mine is from
September 13th uh of last year it’s a wedes day around noon and we’re looking for doors numerous leites and interviews mostly side comment have led us to the back end of Amazon’s Logistics the nud a marginalized neighborhood next to dortmund’s logistical Park bestest F mostly known for its bad reputation and clinging
Stigmas sometimes also known for in its importance in BVB soccer history and its racist cops the stories are speaking from the posters and stickers on the wall we’re looking for the doors T literal and metaphorical gateways for gatekeeping practices and old V the entryways to the heart and bodily work
Done at the hsh factory for example our interview Partners kept describing the nud as a Gateway itself a neighborhood of arrival Amazon as a door opening infrastructure from a work perspective is an interesting but controversial point although the German working conditions are precarious exploitative and temporal Amazon’s jobs are relatively easy
Gateways to a get a job within having to speak without having to speak German especially because it’s english-based and thus be ensure a temporary stay with the country within the country which makes it very attractive to migrant Workforce who depends on easy accessible jobs for their visa
Processes albate it is important to note that Amazon might be a door opener but it’s similarly a gatekeeper with its culture of higher and fire extremely hard bodily work in the warehouses and the always lingering threat that workers might be fired after their probationary period and or when they’re politically
Active in a union for example Amazon is strictly controlling their workers building on an extreme toxic dependency on their jobs within the more vulnerable group of of Migrant workers ironically also based on the language and local legal knowledge is a do at Amazon consequently part of a social migratory
Infrastructure in a neighborhood such as the nstat or in Hamburg the feder what is not merely enough talked about in research about tech uh and especially Amazon we’ve noticed is what these infrastructures that we’ve been talking about in the in the past um what they have H what kind of impact they
Have on urban areas and urban ways of life but if you look closely and follow its logistical traces on your street in front of your door or your favorite kiosk um you come across an intricate web of Amazon infrastructure in this Amazon town oh hold on the slides aren’t
Working want there you go in this Amazon town the Halo This Little Dot that sleeps next to your um bed is an alarm clock it tells you about the weather that is outside it also tracks your Vital Signs and how you’ve been sleeping while you are waking up with
The Halo Alexa orders your coffee already um and also your Uber and while you’re going to work you come across quite a few delivery winds or even robots once at work as we’ve already mentioned you lock into your mostly and probably AWS Based Services and interfaces on your way back home you
Kind of pick up your your uh Amazon uh package at an Amazon Locker that is maybe at your favorite kiosk around the corner and when you arrive at home um your home awaits you with uh heated um um with the right temperature with your recommended playlist for the evening and
Maybe also a filled fridge because Alexa reminded you to go to the grocery shopping um as one of our colleagues Ellison Powell said within an uh field trip that we took together the Amazon infrastructure is quietly but maybe parasitically moving into our homes into our cities taking up spaces and taking
Over already existing infrastructures so the questions that we have here is this logistical City that Amazon is creating um who’s this run by and who’s this run for so in the next couple of slides I kind of want to um draw a picture of what I call uh the urban back ends and
The urban front ends um what I’ve already been mentioning and Army and no have already shown in uh the particular ways um so I would like to start with the urban back ends and ironically the urban back end is where the end of the supply chain begins um and we’ve seen
This in my fil note um those are mostly or the distribution centers that we’re talking about um when we talk about the urban back ends and the start of the logistical process that goes with into the City and delivers the packages to our doorsteps um is mostly based in
Neighborhoods that are on the periphery of uh Urban centers um or when they’re in close to Urban centers or in the urban center even they’re mostly marginalized and we’ve seen this in the NAD in Dortmund but we’ve also seen this at the feder in in Hamburg where I am based so these
Peripheries um have often the not only the problem that they’re marginalized in many ways they don’t have enough infrastructure it’s very hard to get there it’s also very hard to get to work there um when you’re working for Amazon or if you’re delivering for Amazon they’re also very much um affected by
The pollution of these industries that B that Amazon is one of and they also um are the place where this very bodily very hard work of the worker happens um and um at the same time we have to say that also the workers not only work there but mostly also live there um
And are yeah it’s important to say that um these neighborhoods thus also kind of operate as Gates and we see one of these Gates here um metaphorically um as gates for a migratory Community where um in within these neighborhoods are already existing infrastructures to um
Kind of get work uh in Germany get visas there’s knowledge circulating around um ways of living and everyday life jobs are passed um and recommended um and thus uh the back end of the Amazon infrastructure is basically buils and it’s often invisibilized um to the customer since
It’s at the periphery um and we as customers or if you’re a customer mostly don’t see uh where this delivery process starts here we can see the crowded Subway um that these back ends often hold and a worker going to work also what we can see is that Amazon is
Mostly renting spaces they don’t build their spaces and what is interesting is that they temporarily take over these spaces and then maybe also all of the sudden vanish again um so it’s not very clear for how long the contracts are running it’s not very transparent um who
Owns these real estate properties um and uh yeah we only see that public infrastructure like the spaces um the congested streets the occupied sidewalks Al also in an algorithmic way uh through Amazon sidewalk are um taken up are taken over um there’s a lot of parking
Lots um and lockers that we see within the S uh in the city and um increasingly not in Germany but in the US you’re probably very familiar with the Amazon supermarkets so space is a really important bait for Amazon to come to a specific region or
City um and even though it is rare cities go out of their way to attract Amazon to their communities um this goes as much as we’ve kind of already mentioned um this goes as much as giving out public data um or Civic data even of the communities uh in the application processes for
Example for headquarter cities and I want to quote danan here um who says Amazon got all of the city planning data on the future construction of America for free as well as bidding information that offered insight as to who might Pony up fed subsidies that was
Always the goal for the hq2 project and he’s referencing uh their second headquarter project in Arlington that was opened last year so this takes me to the urban front end and the shiny surfaces of not only privatization but the big headquar buildings that we see uh coming to
Cities such as Arlington and in Germany this would be Berlin um where they have also or are also in the process of opening up a second headquarter uh for the germane speaking market and Country um ironically these front ends are for the workers who work on the back
End so for the um developers the programmers um the people mostly that work at AWS so there’s kind of a reverse logic to what we see in spaces um but as I said um Amazon is very much profiting off of free public data and subsidies while um being
Attracted to this community and you know uh being embedded Within These communities so we also see a parasiting um or parasitic structure um taking over you know neighborhoods um while maybe gentrifying because they bring specific a specific um white color Workforce to these neighborhoods um and we also see
That in Germany and there’s big protest against these uh tendency to take over um spaces and neighborhoods that have previously been very diverse and are now um attractive for tech companies that you know very much bring a very homogeneous uh group of people uh to these uh cities and um
Neighborhoods of course they’re also attracted by knowledge clusters we see this in Germany with universities and research centers Amazon seems to be very close by to a lot of universities and research centers that Focus around robotics K AI um and other um interesting um sites for them and their
Infrastructures um but it’s also um very important to look at the geopolitical Notions of where they’re setting so in Arlington this is very close to Washington DC um I don’t have to tell you that um so this is a very geopolitically heavy uh um location um Berlin for Germany obviously is the head
Um is not only headquarters for Amazon but also um the biggest city um and where politics happen so they also use space to be near these infrastructures to be near the points where politics are decided upon um to be near to people who have a lot of say in who is shaping
Spaces who’s shaping societies and politics so these moo surfaces hide um these gentrification processes that I’ve been talking about um and and Amazon is very much comfortable and settling in these cities also through private Public Partnerships that I uh built um um also privatizing um of you know specific uh
Parks or other spaces uh that we see um and kicking off infrastructure projects interestingly like the building of Highways the building of bridges and Metro stations and another tendency that we see um Behind these uh shiny surfaces is that cities are increasingly used as test bets where Technologies are
Prototyped um where things are tried out where infrastructure that that is owned by Amazon like their supermarkets and Laboratories and others are implemented uh within the neighborhood tried out um as sort of like a um a real life lab in a way so talking about these spatial
Injustices that I’ve been pointing to um the question arises what does this do to a city um and who does it does that to basically um what happens to the exciting density of cities the exciting heterogenity of cities where people of different backgrounds get together negotiate the urban um and you know
Build Futures that I diverse plural um and when we look at the Amazon town that I’ve been painting a picture of who’s represented Within These spaces and within the technologies that these spaces um hold who can take part in these infrastructures what we see is that Amazon is very much trying to um
Establish a controlled and curated public um and does this also through means of convenience and very very smooth Logistics that we barely notice and who uh which vanish in the backgrounds of our conscious basically this is a little um yeah poster from the um from the uh how do you say I don’t
Know like a movement against Amazon in Berlin it says safe your Keat which the Keats is like the neighborhood and it’s fighting Amazon basically so the end of the supply chain is just the beginning um that’s what we’re arguing um when we look at the end of supply chain um which
Supposedly could be our home um our home itself is again feeding data into the technologies that our home assistants like echo or Halo that I mentioned before um so really we’re not talking about a linear uh supply chain we’re mostly talking about a circular supply chain because the data is fed back into
The processes that for example ol mentioned the algorithms um it’s learning um it’s um optimizing these algorithms and you know algorithmic architectures which go back into what Armen was mentioning the logistics on the ground uh and the smooth um flows that are Logistics and then come to our home and are going back
Into um what we call our everyday lives and then the cycle pretty much starts again um this is just a little reference to the um music group um f lisbet it’s a German band who says Alexa please give me back my money um because obviously it’s not uh holding what they
Promise so we what we want to make clear um to sum it up is is that it’s very important to um critically question these infrastructures that we’ve been talking about um but especially the role of big tech companies not only within you know technology studies within um media studies or logistical studies but
Also what impact they have on Spaces because space is essential to our everyday lives um especially in cities um and space represents uh what social um implications we negotiate and um live by so in the light of public infrastructure being more and more and often invisibly influenced by capitalist
Tech companies such as Amazon we need to continuously interrogate their authorship standards accessibility and influence on Urban everyday life in space the Amazon ification of logistics involves standardized labor regimes with heavy algorithmic integration and soft cultural Management unionization in delivery is far away collective bargaining promising Justice even
Further and with that I’d like to close and also say the wrong Amazon is burning um we as researchers feel um or as researchers Amazon especially feel like we also have the responsibility to keep the conversation going to not only open the black box but make it accessible to not only an academic
Public but a public in general um so yes we want to say um have a lookout and maybe question um what comes into your home and what you see on the streets and fight for what you think a future city should look like and be like with or without
Technology thank you so much much uh for all of you for that uh I think amazing uh presentation uh of you know some really you know big work right I mean I you know wide- ranging work work that’s you know uh that feels very large and significant to me I guess the downside
Of that is that the target of your inquiry is also very large and you know uh unwieldy in some ways um but uh does anyone um uh have a question from the uh audience that they’d like to ask I have a couple questions but uh uh do you wna
Oh yeah here I was gonna ask you there we go if you would stop sharing your screen Miley but uh I can make it happen uh so we do have a question uh from IP IE Zoom I’m gonna uh allow you to ask your question go ahead
There uh can you hear me yes I’m can’t see myself okay perfect sorry about the the zoom name this is what happens when you sign in on multiple accounts and don’t know which one you’re on uh my name is PR it’s good to see you again am
And thank you for again for presenting this incredible work um I just had two sort of points of clarification and sort of followup questions uh one is the uh what all brought up with regard to the U sort of traveling salesman problem and how that’s been extended in some of the
Fleet Management s of technical literature into the vehicle rooting problem and specifically how that has been that has found its way sort of as uh to address the vehicle looting problem through Google’s or tools there’s some interesting um references there in the OR tools literature specifically about how the distance metric is calculated
And relevant to that um some of the efforts to sort of sort through what you call the messiness of um these roots the the root data or the rooting data has in in some ways I think in an interesting way being handed over to the um to the drivers themselves
To make what some of these for example apps called rooting choices as in they allow the driver to like the the driver is given a number of choices about an optimal or efficient route to take to get from point A to point B um and then once they make that choice that then
Gets fed back in so I think that’s an interesting point where the sort of apparent choic of agency given to the driver in making the sort of last mile delivery possible is then fed back into the aloric sort of systems and the Machine learning models that then again
Optimize for roots and in fact reduce the number of choices um so I thought that was interesting the other thing I wanted to uh sort of talk about or like ask about in the towards the end um in my’s presentation about the sort of the overarching theme about you know
Amazon’s effect on public infrastructure or public space and sort of urban space um thinking also about how the sort of historical making of the city of certain spaces has allowed or made it easier for companies like Amazon and other logistical sort of companies and supply chain companies to
Um enter these spaces and remake them in ways that are uh conducive to do uh some of the sort of Supply Chain management in cap forms of capitalism um how we sort of with that knowledge how might we um because there was a note I think at the very end about
How the difficulties of sort of unionization collective bargaining which are well noted but I think there are also aspects of that that have already resisted some of the same um you know moves that companies previous to Amazon have made in order to try to remake Urban space and how those have been
Resisted including say anti-gentrification efforts but a whole lot of other sort of tactics and strategies of the assistance and how we might look to those with the understanding that uh they can be applicable to resist a change how um Amazon is attempting to remake the city so yeah thank
You I’m not sure if I understood your question about the travel Le salesperson problem right perhaps before that the O tools um by Google this is a um set of algorithms that um uses the ristics that are well known to get an really almost optimal result and this could be used by
Everyone because they are open source that that already answer the second half part of the first question uh sorry just to clarify I just wanted to note that uh the sort of as you noted yes the tools but I think the there is a there’s an interesting sort of extension
Of those of the traveling salesman problem in the O tools’s approach to the vehicle looting problem which is more specific to Fleet Management that I thought might be just interesting in order to study how some of these roots are constructed based on the decisions all like the live decisions of the uh drivers
Themselves yeah send us the material we have a look can I answer the middle question and then Maya Le can maybe try the last one so this the stuff about routing and what the drivers are allowed to do not is kind of interesting because um like
You know like I was even re looking at some of the material just to clarify because it’s fascinating that the ring challenge says here you know here’s data on how drivers were driving and clearly not according to an algorithm and you try to recreate that and we read that
Basically uh you know these drivers doing data work that then gets extracted and put into algorithms so the algorithms become more optimal and not in the way you know mathematical models would have you know suggest optimal Roots but taking account of kind of extracting that knowledge local that
That knowledge that is local embodied and so on quite sure really if there’s a kind of a um kind of a straight a strict kind of one answer to the question of you know what’s the relation between how much Liberty these drivers have my my assumption would be that you know if
They do well they don’t get punished and if they if if they are too slow they get into trouble but in any case they’re doing the data work to feed the algorithm to to you know to improve the optimization yeah thank you so much for your comment and question um I think
They’re very interesting and we’re still trying to tackle those questions as well and finding answers and as you might know like uh research is always uh starting with questions and with even more so I don’t know if I have like one exact answer to your question but um concerning the capitalist cities um
Where companies have or seem to have more power these days um I think I have to clarify that it is still very different between US cities and German cities I’ve seen this um in on a research trip that I took last year in the US and I feel like uh companies have
Even more power that’s just from an outside View uh in the US and um Regional and City politics uh that was what I saw on the surface um and also a higher acceptance um we have to say that when we talk about Amazon uh interfering in cities here um we’re question a lot
Because the city is very respected as um an institution to run things and so our regional politics um um or even the state um that provides most of our essential infrastructure still so I just want to point out that there are very different um perceptions of what I would
Just you know casually call the capitalist city and what companies uh can have an influence on those um and in terms of what to do um resisting these structures um that’s also a very tricky question question and I think what is essential in what we’ve seen in New York
City for example with people rising up against uh the second headquarters being built there is that um looking in the past as you said is super important also exchanging knowledge you know generating knowledge around what it does to a city when a company like this um comes to
Your city it comes to your neighborhood um builds a headquarter there um I think that is very important so collectively connecting each other um building um alliances um around the world and I’m Happ to say that I see that globally right now but it’s also important to say
That as a citizen of your city it’s you can you have a say in what is done so you can go vote um that’s like the most basic one if you’re able to vote um which is not always the case so this is also tricky um but you can go to town
Hall meetings you can you know uh comment on the poit politics that are done and I do see that that takes a lot of of time and also energy and capacities that not everybody has so this is not very you know uh easily done
Um but it is a way to kind of get involved and to be critical and to ask questions um to you know what politics are doing and who are they cooperating with and what impact uh tech companies have within your neighborhood or city yeah I think um reading arman’s
Comment in the chat I think Yeah tomorrow’s uh talk is is very on point here in some ways thinking about these you know histories of resistance and how we can incorporate them and bring them into the discussions we’re having now um fortunately I guess you know respecting
The city is not a problem we have here in the United States in terms of thinking about how a city like New York is handling some of this stuff but um it you know the vehicle routing problem you know makes me think of the sort of early
Um kind of UPS uh algorithm uh which was so incredibly simplistic right which was that drivers cannot make left turns so that it would all it would never route people to make a left turn because you know on a right turn you could potentially turn during a light um you
Know driving the United States uh and to go from that to the kind of level of complexity we have now I mean people you know thinking about the public perception of how these systems work people can understand oh yes I drive I you know it’s easier to make a right
Turn sometimes and a left turn I can understand that algorithm but then you go to the yeah sort of level of complexity that these algorithms have now or will have into the future um and you know it’s almost completely inaccessible to a sort of average person
You know how these things might work um are there any other questions from the audience I have some questions um so I’ll I’ll ask a question if anyone would like to ask one go ahead and raise your hand but uh I’ll I’ll take uh prior it
Here uh because I want to you know think about your work in the context of sort of broader um uh broader work in the critical study of logistics um and I think Armen started immediately by saying you know uh you know we found some workarounds because you have to find workarounds and
So my question is about the workarounds you know um it used to be the case I guess think as a historian that some of these companies you know companies like Western Electric and Bell you know or I even IBM they’d be quite happy to tell you you know if you were an academic
Researcher exactly how they do things at a certain point in the past but today you know companies like Amazon and other logistics companies are so closed off and inaccessible to us that we have to do you know all of these interesting workarounds that appeared throughout all three of your talks right thinking about
Simulating possible routing because we can’t really understand you know we we Amazon’s never going to say here’s our system why don’t we route some packages and open that up to us you know Finding where the interface for some of these systems you know Armen you showed you
Know a picture of the interface you said this is from a video you know I found you know we don’t even have access sometimes to what these systems look like right um the flex example is also quite an interesting one because even though Flex has you know uh been sort of
Discontinued or you know my local experience is it’s been radically contracted in you know the sort of amount of use that it it has um but we have you know the flex interface from when it was super available that interface you know how some of that
System works is still in the hands of you know Amazon’s you know Delivery Service Partners right because they’re in some sense using the same system so that is yet another sort of workaround and inroad so just you know kind of reflecting on your work process so far
You know what um you know and of course those I think you said four four in you know four official interviews that you got or something like that right um kind of on the record interviews with Amazon might contribute some help also but you know what have been the sort of most
Valuable workaround strategies I guess is it just looking through corporate documentation finding you know here’s a screenshot in a video is it doing this kind of speculative design work um thinking about how to simulate how these processes might work you know what is your take on you know other researchers
Who are looking to study these kinds of systems in terms of the workarounds that we all have to employ should I start or do you want to one of you I mean my I my my quick response would be um you kind of need all of them because all all of them give
You very partial knowledge I mean just to you to raise the stakes on this it’s a really difficult problem because even you know the Baseline you would assume would be that you know Amazon also employs loads of um Scholars and scientists right and you think they talk
To you they have a science blog and you think you know this is science the tsp problem is a science problem right we’re we you know we’re Scholars you know we Humanity Scholars but we’re you know we’re Scholars and they will talk to us you’d assume and we you mean we’ve
Written to a few and spoke to a colleague yesterday from from Zan who works on Alexa he’s written to dozens and dozens of them and you always get the same response which is you know we’ve signed an NDA and all we can tell you is what already in in the um kind of
In the Amazon PR announcement on the blog and then that that’s the end of that conversation so it’s really like even at that level you would have thought that would be an obvious way to kind of back door to you know to do what we do best which is you know study what
Other Scholars do and understand what they’re doing um so so that door is also shut and I think that was probably very different earlier um but I think you need all I mean I I don’t really have um we’ve gained this is the exciting thing from all of these things we we’ve
Gathered really the patent research think that for example we had a master student Clara Frieza talking about kind of emotion recognition at Amazon in in voic assist in voice you know voice assistant and you know going through the patterns and what they write about what emotion means to them like how how else
Are you going to find that out so like you’re right we have this very big moving Target and we go at it from all different directions but I didn’t really like there’s no not really a a royal road all of these workarounds like and it’s still like sometimes the pieces
Fall come together sometimes it remains quite patchy like with this question of like you know do they follow the routing or not so I I don’t really know I don’t know if any the others two have a suggest but I don’t really find that there any of them will one of them will
Do I think the key is um to accept that we’re not going to open the black box but we’re working around it so kind of thinking outside the box quite literally in our case outside of the Amazon box is essential and way less frustrating to try to kind of figure out what the
Inside are so we’ve kind of embraced that feeling after I don’t know one and a half years of research and said okay we’re probably not going to have a very clear look inside uh let’s see how we can manage around it and you know just going to the neighborhoods that um you
Know Amazon is operating in um trying to talk to as many people as we can um also on the streets have been and kind of like focusing on the everyday life of of these infrastructures and the logistics um that we see uh has been extreme extremely revealing and helping and uh
We’re also very grateful for the people who wanted to talk to us but um what is tricky is to also protect these people as research because we do have a big responsibility um when we get interviews and um they have to be very careful as Armen said you know they’re uh you know
Followed for talking to researchers and they’re very suspicious and um all that out of understandable reasons so um it’s kind of hard to paint that picture without um threaten or without someone being threatened by our research um so that’s always an important thing to keep in mind um considering ethics and all
That yeah um Eliza wrote in the chat um the cynicism problem for the average person means that a vast majority of citizens either don’t know the extent to which these companies are endangering our lives and well-being or don’t care have a kind of sense of resignation or nihilism how can we work to
Realistically provide people with tools of resistance so that these conversations aren’t just limited to academics researchers um you know Etc who already agree with each other and yeah I think this is this is an important question I note that we are almost at time uh so I wanted to read
That into the record I don’t know if any of you want to respond briefly to that um I can and just to point out like I mean there’s there’s different different things like the union is actually very good onto the job now right there’s like a internationally organized kind of unions that are
Sticking together and and plotting about how to how to deal with kind of the labor issue specifically and the other thing that was for important for our project was to bring this to the city to to not this is there’s a Terrain here that you know where this stuff happens
Where you can get active so so on the one hand you know yesterday myi organized this um Workshop where you know we had 30 people talking about these Technologies and how they impact our daily lives a lot of this is like you know in the register I don’t know if
You know things like the tech workers Collective and the work that my My Le and her group do this is like you know on the ground education around these kinds of how this Tech impacts you know our Smart Homes for example but also the other the other ways talking to cities I
Mean you know this is the terrain the city governments you know these are one of I don’t know you know in the US um in Germany municipalism is kind of doing okay as kind of like there’s lots of messiness but but also there’s people to talk to this this this like Democratic
Structures and and you know just just pointing out that Amazon is an actor remaking our cities and having starting conversations about this with the city with the municipalities um that that that’s been somewhat refreshing that’s where we’re pushing a bit as well um because you know they have they have
Scoped to deal with this and they’ve noticed you know they’ they’ve all been kind of I wouldn’t say count by Amazon but they’ve given them all of these big spaces for their Logistics fulfillment sentes and then they realize well there’s there’s there’s very little jobs
For the amount of space we given and now we have no space they’ve taken up all the space what do we do now you know um on the labor question but also otherwise you know like so they’re they’re kind of starting those conversations as well and so that’s also and there’s probably
Other spaces we could point to that or that you could think of that that I were taking these conversations to yeah um yeah lifting up activist work Miley says in an academic context is very important building alliances which is yeah something that we we hope the the research Network you know continues
To build connections between uh activists and organizers you know again thinking about that question of workarounds um even if you weren’t going into studying you know some of these supply chain companies and processes from a kind of activist uh inclination the resistance you encounter almost forces
You in that direction I think as an academic or at least that’s my experience not to put you know words in your mouth but um uh I just like to thank you all again since we are at time exactly uh 1:30 here now for that you know truly wonderful presentation I
Think you know thinking about the context of your work against uh you know the the previous presentation by you know Jess beer and Jessica Steinman I think he has all these thoughts in my head that I didn’t get to ask about you know transparency and internal transparency versus
External transparency and you know how is Amazon a digital company that is you know applying digital methods in the in the physical world you know there are all these I think interesting parallels uh that we can draw across our uh talk so far uh so thank you again so much uh panelist
And uh thank you everyone who attended uh we really appreciated it