welcome everyone my name is caroline jones and i'm here representing the transmedia storytelling initiative in mit school of architecture and planning tsi is the producer of tonight's event 3d5g surveillance and agency and i first want to thank dean hosham sarkis for hatching the idea for the transmedia initiative and to funders nina and david fialco for giving it wings the small transmedia team began thinking about this event last fall when my colleague in the department of architecture sheila kennedy sent me a seductively produced video advertisement showing the coming of 5g broadband service to metropolitan new york we both sense the immediate crux represented by the convergence of 3d and 5g technologies what do we mean by this drone hoisted 3d spatial video and technologies of photogrammetry both featured in that 5g add and pervasive tools in architecture are already here and when these spatial technologies meet 5g broadband users gain the capacity to process the streaming data from aerial surveillance in real time will publics find this tool useful for their own community monitoring or will these capacities produce yet more centralized pooling of our data this time harvested from above such questions are compelling for mit as a land-grant university born from the technical needs of an industrializing nation eager to cartographically and demographically capture its settler and indigenous populations although we're meeting virtually tonight i would like to speak the words of mit's recently achieved land acknowledgement which is the first step in a much needed process of truth telling about the resources appropriated to produce this thriving institution so here we go mit acknowledges indigenous peoples as the traditional stewards of this land recognizing the enduring relationship that exists between them and their traditional territories the land on which we gather is the traditional unseated territory of the wampanoag nation and other language groups we acknowledge the painful history of genocide and forced occupation of their territory and we honor and respect the many indigenous diverse peoples connected to this land from time immemorial mit is ready to do this work of an acknowledgement and critique and i hope that 3d5g can play a small role in revealing the tools we must wield as engaged citizens for example as we gathered to brainstorm this event we reached out to our colleagues in the social and ethical responsibilities and computing group working on the onboarding college of computing here at mit they suggested aero astro robotics engineer luca carlone whom we are so pleased to include in this discussion tonight with architects nicholas de moncho sheila kennedy and brad samuels we included in that brainstorming my colleague sarah williams and she kindly agreed to be tonight's moderator she'll be providing more fulsome introductions to our panelists and here i need to update you all about the schedule of events at the last minute alas we lost the virtual presence of artist lawrence abu humdan once he realized the time difference approximately 10 hours later this new parent of an eight week old baby understandably begged off however lawrence has made a quite brilliant and difficult video about his own pioneering of forensic acoustics and we'll be streaming that video at the end of tonight's conversation taking lawrence's slot in the program we are thrilled to include the practicing architect brad samuels of the firm situ again sarah williams is going to provide much better introductions to our panelists but suffice it to say that the work of situ in mapping drone activity for the united nations showing us hidden new york as well as conducting forensics of police encounters with protesters provide much to discuss for our later conversations it was generally a boon to have sarah williams thinking with us as her own research in the department of urban studies and planning here at mit centers on marshalling data from public records to contribute to community empowerment director of the civic data design lab and now newly leading the leventhal center for advanced urbanism sarah has recently published her book data action with mit press focusing on harvesting and arranging data in compelling forms to show urban inequities such as rent burden or prisoner re-entry and work patterns in the city before i met her i encountered her work as part of the collaboration the million dollar block project which she participated in while still a researcher and a grad student at columbia i saw this work at the museum of modern arts exhibition design in the elastic mind to summarize its point the project showed blocks in brooklyn in which the state was spending a million dollars to incarcerate several inhabitants from that single block the data rich graphics were arranged to ask the counterfactual question what would happen if each of those million dollars had been invested into education and social support of those citizens before they were incarcerated continuing her focus on equity professor williams civic data lab has recently produced a project utilizing data from greater new york given to citizens entitled visualize new york commissioned by the center for architecture and opening in november 2020 the virtual exhibition and website continues to invite new yorkers to think about their urban realities in the run-up to an unprecedented election scheduled for november twenty twenty one sarah's lab organized data visualizations across five key topics climate change housing streets and the public realm neighborhood health all with a lens on equity so without further ado i want to hand over tonight's program to sarah williams and she will introduce the panelists and set us up for a lively discussion on the civic implications of 3d 5g um thanks so much caroline um for that wonderful introduction i would also just say i didn't go to grad school at columbia in fact i went to grad school at mit uh but uh just thanks you were brought into that project somewhere in the country yeah i was the co-director of that research project um in the spatial information design lab i was the co-director of the spatial information design lab while i was also a faculty member there in the planning department before i arrived back at mit um so i am just going to go share my screen um for everyone um [Music] so i'm here to give us a little bit of a primer on what 5g and does hopefully you guys all know what 3d is already um but the hype around 5g surrounds its ability to process unprecedented amounts of data over cell phone networks allowing for new technologies such as autonomous vehicles which depend on the ability to process huge amounts of data in real time the create data created by these vehicles produces a recreation of the world every second and 5g it creates the ability to send this data back in the cloud real time here we're looking at this point cloud data coming from google's autonomous vehicles um and you can just imagine the amount of data that every car creates on the roadways um french cd artist antoine delash 2016 piece critiques this ability to scan our world in real time in a project called ghost cells this six minute short creates eerie almost apocalyptic paris and it won uh the audience prize at the claremont friend international film festival and while it's thought that 5g combined with the data that can be connected through drones has the potential to do a lot of good in the world such as this 3d reconstruction of italy post-earthquake or this example of remote surgery as seen in china all facilitated over a 5g network there are many potential harms such as baltimore's persistent surveillance system this is one of our aircraft it's the one we're using here in baltimore this is our camera system it's 192 million pixels which makes it about equivalent to 800 video cameras at once and everything is fed down to the ground live through our data links here we do wide area surveillance we watch cities at a time in order to assist law enforcement and lowering crime rates we're essentially a google earth live with tivo capability that allows us to follow people to and from crime scenes here's how it works ross and his team listen to emergency calls and police standards to find the time and place of recorded crimes they go to the live feed that their plane is sending down and start to scroll back and forth through time to actually see the crime take place so right here you can actually see two cards that match the description from the 911 calls the cars leave the scene here and we're going to try to follow those cars and see if we can find a final location that they go to and we'll also follow them backwards in time to see if we can figure out where they came from this scary technology which is also used to create 3d representations of boston um and has been facilitated by the 5g network has now been twice shut down by the aclu who has pursued legal recourse to take these systems offline the first time in 2016 and then more recently in 2020 and as 5g comes online and the data can be more more easily passed to the cloud i would argue that we need regulations for how these devices are used and while baltimore's project has been shut down just recently um in places like kansas city they are already deploying these drones to do the same thing right now regulations for new technologies often lag far behind take for example gps tracking and while gps technology has been around and widely available since the early 90s using them to track people in a criminal investigation just recently became illegal without a search warrant i would argue that we really need to consider all the potential both good and bad that 5g technology offers us and how we as architects designers and engineers urban planners uh use these system and what ethics do we bring to bear what potentials do they hold for us so hopefully i will pass it over to our amazing speakers to enlighten us on some of these issues first we have john dimancho john dimancho is a professor and head of the architecture department at mit as well as partner in the architectural practice modem he is author of spacesuit fashioning apollo um which was released by mit press in 2011 and local code 3659 proposals about data design and nature of cities uh by princeton architectural press and 2016. sheila kennedy is an american architect innovator and educator she is a professor of architecture at mit and a founding principal of kva maddox an interdisciplinary practice that is recognized for the innovation architectural material research and the design of new resilient infrastructure for emerging public need her work in design research examines intersections between natural con ecologies and hybrid high and low technologies and network cities and urbanizing global regions impacted by climate change we have luco carlon assistant professor in the department of aeronautics and astronautics um dr carlone directs the mit spark lab which works at the cutting edge of robotics and autonomous systems research his research into geometric computer vision applied to sensing perception and decision making in robot systems yields expertise at the confluence of 3d and 5g brad samuels is a founding partner at situ and the director of c2 research an organization that merges data and design to create new pathways for justice outside the multi-disciplinary practice brad sits on the technology advisory board for the international criminal court the advisory board for the carnegie mellon center of human rights science the advisory board for dartmouth's rice center for the study of computation and just communities and the board of the architecture league of new york he is a fellow with the urban design foreman and teaches in barnard college and columbia university undergraduate architecture program thanks so much to all of you for joining us today i will pass it over to nicholas dimancho thank you so much thanks so much uh sarah very happy to be here with you all tonight i'm looking at the list of names i think they i would emphasize that they probably go in ascending uh potentially ascending order of actual expertise on uh technologies of 5g networking uh um as well as drone aviation and and all the rest uh my own background in recent years i came to mit from uc berkeley where i was for several years director of the berkeley center for new media uh where we thought and and and still think a lot interdisciplinary about interdisciplinarily about the nature of uh contemporary technologies and their effects on um uh ideas uh and even the university itself so with that i'll explain the perspective of uh my provocation as i was asked to uh uh to share with you and then um uh we will uh we will take it from there um yes okay i'm getting a thumbs up so this is a colorized version of exactly the same class of image that sarah was showing you a google autonomous vehicle point cloud um and the uh again to emphasize my own uh take on this i'm not so much an expert in the underlying technologies but certainly our participant in a whole set of conversations over the years thinking about the relationship between information and the city and society the two are of course inextricable and uh with uh with a slight um uh warning um of uh the probably the most disturbing image i'm gonna show uh this evening which is the this image of the recent um uh capitol hill uh uh insurrection is one that can't be far from the mind of anyone really thinking about the intersection of contemporary data and society and physical space this is of course a different kind of conflation than one we're talking about but not unrelated in that it's a on the one hand a physical manifestation of a massive online data-driven phenomenon of social media the tendencies and innate proclivities of data-driven media to amplify difference and uh and and the physical space of the city in this case in a particularly violent and disturbing way uh we see on the screen here thomas fee of freeport new york he himself was arrested later rightfully so because he texted this image to his brother-in-law a federal agent uh revealing the fact that we also still somehow think what happens in virtual space and physical space aren't related and yet our world is ever more defined by that intersection and the intersection of it with physical society uh it's a uh of course coming off of uh uh the uh uh sedition uh uh and the the capitol hill uh uh invasion of data uh in on january 6 was a massive uh was a 140 000 image data set uh created put together by the federal bureau of investigation and a one-day 26 spike in the use of the uh app clearview now very well known little startup that has scraped three billion images from facebook and instagram and linkedin and others um uh reminding us of course of the old saw of the study of social media in its early ages which is that if you aren't paying for a service you are the product [Music] so from that slightly tendentious uh initial thought comes a couple of subsequent thoughts about the nature of the relationship between data and cities themselves and what this massive increase of information is likely to mean in terms of a massive increase of all kinds of urban phenomena and civic phenomena that we associate with um uh the proliferation of data and data about us in particular um this is uh an uh an image i like of uh what was the the nascent but now no longer anything at all effort by google's alphabets division uh sidewalk labs to build a massive new neighborhood in toronto this is the sort of visitor center and behind it is probably a not coincidentally placed um uh apple billboard that says if you can't read it we're in the business of staying out of yours uh one of the the the massive investment by um uh uh by sidewalk labs in in the end a fairly you know inoffensively innovative uh architectural proposal was ultimately undercut in the local politics of toronto uh by by a lack of trust in um uh google's data handling of the data produced by residents of this new development which would have a massive 5g enabled data collection would have had a massive 5g enabled data collection uh infrastructure designed to improve quality of life uh leading to the sort of civic data trusts and ideas that were you know in the end unconvincing of course that's all a bit distracting because uh what we're not probably at risk of is massive google-built uh housing developments in our cities but rather at the convergence of social media 3d technology 5g networking in conflating in a less dramatic and alarming way the space of media of social media data and media about us and the physical fabric of the city in fact in an even less dramatic way it's already happening with things that you're not likely to have opted out on of um when you log into certain apps on your mobile phone the the last thought that that i would bring you with to zoom way out but hopefully before all the rest of our speakers zoom way back in is what i think should be a particularly interesting concern um but but no less urgent when it comes to this uh geometric proliferation of data and our own experience of the city which is its threat to evolve things anonymity not that i wish the invaders of the us capital to be anonymous but the history of the city and its mode and history of its um identity as a space for innovation has relied extensively on it i think it's worth pointing out um this is a famous essay by evgeny masorov lamenting the end of the so-called cyber flaner the age of the internet ending in about 2012 when it was all about wandering around places that didn't know who you were and exactly what you were doing there and this ironically is a very enough is a very different earlier form of media media a typewritten letter by jane jacobs where she makes the case for this kind of anonymity in her own pitch to chad von kilpatrick of the rockefeller foundation for the manuscript that would become the death of life of great american cities where she emphasizes the loose and and and anonymous connections between members of a shared society in the city as one of the sustaining elements of civic space when um uh 19th century sociologists like durkheim and simel wrote about this the the city they connected the anonymity of the city uh to the experience of depression in it but what we've learned from uh urban studies after this fact is in fact in established urban societies in particular mental health is higher in cities and is tied precisely to our ability to take on and act multiple separate identities in multiple contexts as well as to be part of an anonymous whole now of course social media by contrast has been linked to depression for its ability to keep us in a single and a scroll uh or feed of information which has its own of course ambient effects of the city in um uh uh our tendency to to shop online and to have uh effects even on the spaces of sidewalks but for for myself again as a thinker about media in the city first and foremost um i think one of the most hopeful um uh uh and troubling questions is what will a massively increased amount of information about who what and where we are in the context of the city due to the city's unique and essential balance of chaos and order um uh out of it is is both the uh uh the sort of dystopic possibility that these two will become ever more enriched and the rules of social media space will ever more be the rules of the city but also potentially a hopeful prospect that we will start to truly understand and push back at the bargains we've made in one set of spaces when they start to emerge on spaces which we are more likely to see ourselves as civically occupying so that is uh the the the provocation i have to share to start us off and i look forward to handing over to um our next participant thanks thanks so much nicholas a lot to talk about leader now i want to pass it to luca my name is luca carlona and i'm the director of the sensing perception autonomy robot kinetics lab at mit malab is working on developing intelligence and autonomous navigation capabilities for mobile platforms for instance we design algorithms that the drone can use to build a 3d map of the environment in real time using this map the drone can avoid obstacles can perform inspection or support search and rescue operations we also work on the algorithms that enable a self-driving car to identify the traversable space to track other vehicles and recognize traffic lights and traffic signs and in general my students and postdoc work on the design of autonomous navigation capabilities for a broad range of platforms and applications ranging from autonomous exploration with a boston dynamic spot robot as the one that you see on the left to object transportation with the drone as the one that you see on the right so for those of you not working on these topics you must know that we are living in a time of unprecedented progress and opportunities amen aerial systems what we typically call drones are becoming increasingly popular and indeed back in 2016 the number of drones registered with the federal aviation administration was already larger than the number of registered airplanes today we have around 1 million registered drones all in the united states advances in algorithms computing and sensing have allowed the development of drones that are extremely easy to operate for instance is a video from a company i love which is called skydio here the user can simply switch on the drone and have the drone record their favorite training session while flying autonomously and avoiding obstacles and this is a drone that you can buy actually you can buy the newer version of this drone for less than 2000 bucks these drones include multiple cameras and have an onboard computer and thanks to the advances in photogrammetry and ai they're able to build a 3d map in real time and they can detect and track humans in real time with some post-processing they can also build a more accurate 3d reconstruction of the area that they explored now while few research labs and companies such as skydio are developing autonomous drones the truth is that the vast majority of drones are currently operated by human pilots the pilot issues commands to the drone essentially to tell it in which direction it has to move at the same time the drone can send sensor and status information back to the user including image streams or position information and this is where 5g communication comes into play 5g is the 5th generation technology standard for broadband cellular networks it was already introduced in 2016 and started being widely deployed in 2019 5g is designed to enable much faster wi-fi speeds and in particular it allows download speeds up to 10 gigabit per second which is comparably what you get with cable internet this is roughly 20 times faster than the current 4g technology we have on our phones moreover 5g allows lower latencies reducing the time it takes for the signal to be transmitted from the drone to the user and is able to reduce this time from 20 30 milliseconds which is what you would get with 4g to less than 10 milliseconds with 5g this has the potential to allow drones to send more and higher resolution data while also allowing a more frequent control of the drone itself this also means that we can move some of the computation out of the drone and simply just deploy smaller smaller platforms another important aspect of 5g is that it enables more devices to connect to the internet while 5g allowed few thousands of devices per square mile 5g can support millions of devices per square mile well at this point many of you will be thinking why do we need bronze and 5g in the first place and the truth is that the combination of drones and m5g offers tremendous opportunities and i'm trying to list just few of them in this slide 5g can enable a more effective use of drones for disaster response and search and rescue and indeed drones have been already used in recent disasters including the california campfires in 2018 as well as for hurricane response including for hurricane harvey in 2017. in this application drones have the potential to quickly monitor large areas they can create a map to provide situational awareness to first responders they can assess structural damage or even extinguish wildfires drones can be used for the delivery of medical goods for instance delivery companies such as zipline are using drones to deliver frozen plasma medical supply in africa and the united states this drone can deliver a package with up to four pounds within a 50 mile radius in less than 30 minutes and zipline has already delivered over 60 000 units of blood vaccines for measles polio and other diseases and is now looking into delivering carbon 19 vaccines drones also carry the promise of making agriculture more effective they can monitor crops and allow early treatment of pests or pathogens or compensate for inadequate moisture or nutrition preventing yield loss and this is going to be crucial since the demand for crop production is expected to double by 2050. as another application drones can be used for real-time infrastructure inspection ranging from bridges to pipelines and buildings and there are simply countless applications to entertainment from drone cinematography to enabling complex choreographies like the one during the super bowl lifetime show in 2017 where 300 drones were used to recreate the american flag during lady gaga's performance finally 5g can allow fine grain control of the location of the drones flying at any given time potentially enabling the equivalent of the air traffic management system that used for regular airplanes this again is going to be important to ensure that drones are used in a safe way when they become more broadly adopted despite the opportunities we have to remember that drones are a dual use technology not this scenery from smartphone and social media drones can be used for the benefit of society but can also be used to invade privacy and post security threats as an example back in 2018 a drone sighting caused the airport of gatwick in the uk to close for two days resulting in more than 1000 flights being cancelled with more than 140 000 passengers being affected in general if any user can operate a drone we have to be ready to cope with malicious users who can post security threats another concern is about of course privacy and surveillance for instance in 2020 china used drones in face recognition to enforce quarantine during the corona virus outbreak while myanmar has been has recently used surveillance drone for domestic monitoring and oppression finally we have to remember that 5g not only allows connecting drones but can really connect any type of device including our smartwatch smart appliances our computers our cars it is estimated that the average person will produce one gigabyte of personal data per day using 5g and the lack of regulation can lead to leaking personal and sensitive information such as our heart rate our location or the food we eat therefore there is a desperate need for better data governance which can limit violation of privacy and ensure the data is used for the benefit of society while respecting individual freedom of rights whether it's data collected by a drone or by mobile phone we have to remember that data is the new oil and scientists users and policy makers have to come together to ensure that data is extracted processed and distributed in a sustainable way and with the approval of the user thanks so much luca um now i will pass it to sheila thank you sheila thank you sarah very much wow um i'm going to share my screen can you see my screen nods yes great um yeah that was amazing what an amazing set of provocations from nicholas and luca um i uh have the uh the task i was charged by by sarah and caroline to try to think a little bit about um the um kind of material backbone of this infrastructure so i'll be presenting a set of visual notes um that won't weight in so much on good and bad uh possible applications for 3d 5g but rather the infrastructure itself and i'll actually be arguing that although data is uh the new oil it still has a lot of the old oil um um you know i in thinking about 5g and its impacts um uh on on the sort of three-dimensional world um the architectural world um i'd like to begin with with just some language which you see here um with the story of um early electrical technology and how that's being told how is this story being told um one form of agency i guess that the discipline of architecture has um is that it has a history and so it's useful to set 3d 5g in a context um so looking at this kind of um electrical apotheosis um words um like rank vegetation of the coniferous forests that would be the oil on which data is based um we're waiting through the ages there's a sense of kind of an arc of history that is inevitably moving forward um there are a million references to a million uh to to genius and to the kind of um huge uh field of opportunity um that lies ahead and you know there are many examples that i could have picked this one is from deloitte consulting um but it just kind of looks at the the language and the claims also for for 5g you know an autonomous future um talks about smart autonomous vehicles smart ports smart homes and has a lot of promises i think the word promise is used three times in this piece of text um 5g can make it all happen again we have the number one million one million connected devices in one square kilometer um there's a kind of a sense of um a tune tuning here uh in this language not necessarily to civic debate or discussion um but rather kind of attuned to to markets and maybe even investors and that is of course a commonality gas lighting you know in the 1800s the electrical grid in the 1900s and the internet era and 3d 5g share language and claims and um these uh claims are that the technology is inevitable um that it's a march of history and that's a succession of superlatives right that uh gas and then electricity and then and then the internet are the cleanest the most efficient the fastest and always uh these electrically based technologies were framed as being um you know being beneficial especially for the urban bourgeois and it's kind of interesting when we think about drones to look at these large um lighting towers uh and street lighting devices that also are kind of occupying what donna haraway has called the kind of god's eye position so um this is uh actually the kind of god trick as she calls it as donna harway calls it achieving the vantage point she says of the cyclopean self-associated eye of the master subject the dream of a vision without limit it uses its accumulation of information and surveillance as evidence to kind of pride itself ironically on its objectivity so even in the early uh in early european uh urbanism we see the kind of effect of this proto-drone-like flood of light and flood of surveillance and of course um people like ian shaw in his uh chapter on the predator empire and anna fergenbaum from cyborg feminism to drone feminism have kind of made the connection between haraway's god trick and the proliferation of drones um which operates with kind of colonial logic of us in the sky versus them on the ground and the drone kind of performs this logic through its digital world view of targets looking down on people that dismisses ambiguity um and of course um there's always this sense that this technology um you know electricity or even 3d 5g is essential to life it's intimately tied with life with all of our devices and this is an example of a kind of early wearable um i don't know if we should really put that one on uh the smart electric bath with all of its wires plunged into the bath that's probably not that very smart either but examples of the kind of accoutrements and the the cultural and societal embeddedness of these these technologies in ways of life in appliances and then a whole host of kind of other other products so i want to kind of just go out there and and and offer a set of provocations or notes for what this might mean for the kind of agency of architecture and i have several different um ideas about different kinds of design and um there are i think some um force multipliers that one could argue really do differentiate the impacts of 3d 5g factors which are distinctive in this new layering of the technology on all the others and these do complicated the complicate the um the forms of agency of design and architecture with respect to this vision of infrastructure and so i think one of them um is this intensely concentrated incorporation of the 3d 5g innovation ecology um the domination of the kind of entrepreneurial space um by global mega companies that take on quasi-state-like operations um a space in which smaller nimble innovator companies are then acquired by just a very few players so there's an increasing concentration and kind of monopolization and incorporation of this particular infrastructure and because of that i think it's really important to think about this uh in terms of thinking about are our agency um as a renewed investment in in constructing or or reconstructing um forms of collectivity or collective action with respect to this infrastructure so i i've been looking a little bit at um some of the early uh 1980s and 1970s um feminist um uh activist uh uh events uh the way in which they they made symbols like this this coat hanger no coat hanger we all understand what that means um in an effort to kind of reclaim women's bodies this is from houston texas um or the green greenham commons woman peace camp which was a very significant anti-nuclear uh protest and site um for debate and discussion um so uh designed for discourse really trying to use the physical world in every way that we can to bring people together physically whenever we have the chance to do that um and that may be through forms of appropriation like this do-it-yourself guide to remodeling the perimeter fence of a big nuclear compound you have the tools you have a gate you use your imagination and you make a grill for your teapot which is a kind of a campfire around which everyone can kind of meet and locate so i think maybe the charge is to or the question is can we take over every corridor can we take over every piece of circulation and try to make that an open space that's hackable where people could actually meet and gather can we connect every space we can to to nature and to natural systems um can we actually in in architectural design look at the physical linkages maybe circulation becomes much more important ways that we physically can meet and interact with each other um the other provocation is designed for disguise um and that is um the idea of digital twinning um that lucas spoke about this is from the skydio autonomy video on youtube that luca hosted and you see here um a model that was a 3d model a high quality 3d model that was produced by drone by the sky dm videos and their flight paths and the kind of level of detail that you get but it's not just making a map of of one piece of infrastructure or all the pieces of infrastructure in a particular industrial or military complex it's a digital twinning with with a high degree of um embedded information and layers so it's a virtually a whole other city that's happening and we may need to design in as a form of an agency a way to actually hide or protect ourselves from this um not only is it a a digital twin city but it's inhabited by digital people who are strikingly uh realistic and this is a great example from the new york times that shows the kind of fluidity of machine um vision and artificial intelligence working together to produce these seemingly um a real set of characters from like 12 um digital phenotypes um this is the artist hungary who works in berlin this is a recent piece from their instagram and what they do is they disguise themselves um by mingling their countenance with that of insects and bugs and they're known as the bug lady and so the uh surveillance systems of the city of berlin don't really pick up on the bug lady when they're out and about we can also think about the kind of resistance of trees the way in which there's an increase in popularity and do-it-yourself masks and how these masks could be take um natural or leaf-like forms because it's still very difficult for machine vision to uh capture the complexity of a leaf of a tree so that's why we think about the resistance of trees here this is a really interesting project that daryl shinseto an mit smart student did for her thesis she looked at video capture which you can see on the lower right and then began to see what was slightly off frame and kept track of what she calls the s scraps uh from a video feed and daryl's project shows that we ourselves are surveying ourselves with every single video that we take our selfies our our live feeds and so forth and that even from those two-dimensional uh images very realistic point clouds can be made and what daryl is proposing is the physical design of textiles that might complicate um the way in which it might be reissued uh and recycled very quickly but might complicate the the fidelity with which machine vision recognizes it so she imagined a series of these drapes um going over beds and couches and so forth so that um people would be less able to invade your privacy and offer you kind of predictive marketing suggestions and build a profile i'm just changing just a little bit from this kind of adversarial interior to maybe um adversarial exteriors i don't know um but you know it's amazing how many books there are on rand earth and we have to think is this solidity is this earth which actually does uh uh tamper and distort a little bit the electromagnetic waves and the parts of the spectrum where three uh 3d5g operates is this enjoying a new popularity will we actually go to the thickness of earth and kind of revive these ancient materials to try to afford us a little bit of privacy um the third is design for sharing designed for sharing economies um and this has to do again with skydia's uh language uh which is shown here in uh four or five stages uh they call it the arc of autonomy um but i think what uh what what we want to look at is the kind of last stage um which proposes a network of drones that really is not sovereign although we may go there but but quasi-autonomous and able to operate as an infrastructure um that is a service um that people might be able to kind of commission um so um how do we how do we share in such an economy is this even about sharing um in at the mit media lab um has done her in in coding uh project which is inclusive coding and her team discovered that um she could not be seen um by some surveillance systems because machine learning uses synthetic environments as luca demonstrated they're part of this digital twin to a practice and those synthetic environments are flooded with a not a diversity of human beings so there's an incredible amount of bias joy argues in the system and so in order to be seen by the system she actually has to mask herself with with a white uh with a white mask so this is a form of identifying a blind shot a spot in that kind of um social sharing or what is claimed to be a social sharing um you could also think about temporal sharing just in time rooms just in times apartments thinking about plans in architecture not as static but as temporal as needing to change with real-time information like peer-to-peer energy sharing groups of people who are going on vacation you may move to another apartment you may move to another room and we may need to build far less physically because of the kind of sharing um that's possible and then the last provocation i have is designed for caring or infrastructures of care and this really um gets to the kind of question and i think the the imperative to um to create an agency for design that that supports the care of the critical zone itself the critical zone um on earth so if we think about the amount of building debris that gets wasted um this is a giant pile of concrete rubble and we think about the carbon investment in rubble um it's possible that if we treated these rubble with the kind of care that we treat consumer packages and tracked the rubble instead of tracking packages we could have an entirely different architecture because each piece could be tagged as we see here in the work of tim cousin from the mark program at mit each piece of rubble could be tagged could be optimized and could be used to actually create new architectural structures if we had that kind of embedded intelligence and the same is true for our forests we could imagine forest libraries curated forests where uh trees could be digitally replicated through scanning which is already happening today whole trees is a company that is doing this already um and we can imagine a force that's curated so that each tree could be taken out when needed um and could be used for its own form we would not need to invest with um non standardization milling and all of those energetic um expenditures that are associated with that so be a whole other kind of set of renewable materials that might become part or more firmly a part of architectural material culture i want to end with a little bit of the fusion of um the natural the entanglement of the kind of the natural and and 5g the drone uh this is uh a apothecary um in the 1800s and he had the idea that he would attach an aluminum camera to a pigeon um and so we get from noi browner's uh literally bird's eye views some of the first of bird's eye views uh early part of world war one this strange kind of views you know kind of like uh this one you know it's it's like you see these pigeons um and they have their apparatus they they fly about and they don't care about following the person perfectly this is literally a bird's eye view they're going to the written into the roost they don't care if the frame is is constantly tilted or it's not self leveling so i want to um think then about the physical materiality right if you're if you're not a bird and if you are a drone you need a battery um and that turns out to be a fairly significant um battery um there are different kinds of um tear downs online that we can find here's one for the iphone 12 here's the a review of the skydio um s2 drone battery it has something like 20 21 minutes of flying time and it also has a 3 000 milliamp battery which is about 11 watts so sort of taken together if we do the numbers we serve 75 watts here with the wall charger for the skydio batteries there too and then the iphone um that might be attached also to the drone um and if we think about that 75 watts and we go up to the density um that has been discussed of a million devices um per a square kilometer and then we think about the size of new york which is about 300 square kilometers we get to a lot of power um we get to about 22.5 gigawatts of power and the department of energy tells us that that is equivalent to about um over 10 000 wind farms which are based on a utility scale turbine and about 2.4 megawatts it would
be the equivalent of over six million seven hundred fifty homes as of uh 2017 so you'd be able to power a lot if new york city got its promise of a dense uh drone network uh in the arc of autonomy um i'm gonna uh close now talking a little bit about lithium because we were just talking about batteries and about extraction and about infrastructures of care um and so we have to kind of rethink uh how we actually get um uranium uh i'm sorry uranium yes but also lithium in this case um the uh lithium triangle in peru and bolivia um is where 40 of the world's lithium is found it's an extremely extractive process in a drought-ridden area in south america and it uses about five hundred thousand gallons of water um you know per uh you know a very small amount of lithium processed for batteries um so i'm going to show you some images of that process um and i'm just going to stop my share for one brief moment uh while i queue something up um for you and then i'll just close it out this way um i actually um uh yeah okay great i actually dug up on a digital remastering of the song so i thought that might be a good way to kind of end this sorrows come and go [Music] now we see the human race thanks so much shula and thanks for ending us off with some music i'll now turn it to brad um about samuels from c2 studio thanks let me share my screen can you see my screen it says loading yes we can see it okay sorry for the delay okay great thank you um so this will be more of a sort of 5g adjacent presentation as opposed to a 5g explicit one but i think the work we've been talking about today is really going to radically change what we've been doing as an organization we describe what we've what we do c2 research merges data and design to create new pathways for justice and most recently what that's meant for us is really focusing on suppression of dissent protests and excessive use of force usually by police and state actors in public spaces um the kind of main uh sort of source of evidence is photographs and videos and the proliferation of those those assets is just exponentially increasing as we all know a lot of our our work has to do with providing context um so what you see here is the portland police bureau um during a protest uh back in june and part of what we're trying to do is kind of look beyond the frame and think about what's happening and this is where some of the the lidar and the laser scans and the photogrammetry comes into place right you know there's the kind of question of you know what do we what's the context for what's captured in that small vantage there's also the fact that there's just unbelievable amounts of documentation and that's only increasing right and so this is a small snapshot of a protest that took place in the bronx on june 4th that we worked on that i'll show you some more about and it's a you know on the one hand you know citizen documentation uh provides a lot of um evidentiary material to kind of get to the sort of crux of what happened it also provides a huge volume of assets to even try and comb through right but the main goal here is to kind of uh cut through the competing narratives and try and establish a baseline understanding of a complex event the fact that protests are often complex events are often called chaotic events provides a lot of semantic cover for those who actually aren't that interested in accountability right but the point is different tools different methods different techniques can be leveraged to kind of cut through the noise whether that's trying to better understand uh activity of federal forces showing up in portland in in july or it's insurrectionists at the capitol a lot of the same questions same methods same tools and technologies and interestingly for us a lot of our work in this field looking at sort of the relationship between urban space and protests uh was done everywhere but the united states to begin with so that what you just saw was in iraq protesters on a bridge in baghdad being fired at tear gas canisters at low angle or here in hong kong looking at high volumes of tear gas being deployed against protesters this is work that both cases we did with amnesty international but again combining 3d environments photographic information to try and provide a coherent and clear understanding of these events and kind of bring your attention to the core legal or advocacy question and i'm going to talk about one advocacy case and one legal case um the irony is not lost on us everything we've been doing had been everywhere else but the u.s following the murder of george floyd we pivoted really hard uh to take these same tools and apply them not only domestically but to retool them so they didn't take six months or a year to actually do the case work i mean the you know so we could really address the kind of structural and systemic issues that were happening across the united states and so i'm gonna show two cases uh where that work was done domestically again one advocacy uh the other legal and those are distinctions which sometimes productively are blurred a bit about the tools and methods you know sometimes we have the benefit of really high resolution 3d lidar sometimes we don't this is the case in the bronx where we had just open city data so sort of simple massing models of the city but here looking to locate vantage points points of capture of videos documenting the same moment in time right in space time sinking videos the metadata is often unreliable but often we have cctv footage and then we can begin to kind of sync other footage up to it and and figure out what videos might have captured the same moment in time among the thousands that might be available this is a ton of work and just kind of this heavy lifting at the beginning cuts into the time we can actually spend analyzing events but you know this is an example flashbang grenade goes off captured in a number of different videos one of them is time stamped and ground truth we then have a sense of where both locally uh time-wise might occur and also on a global clock so these are the kinds of cues that become very productive um here you know merging and experimenting with how to extract information from these videos this is in charlotte north carolina police firing flash bangs and tear gas against protesters in a very confined space with no egress and nowhere to go taking everything out that we don't want to look at in leaving the model and the part of the video that really is salient in this case in a legal case i'm going to go through two two projects very quickly and i i want to just say at the top is a bit of a warning that some of which you'll see is is violent it's not really intensely graphic but there's some violence that i'm going to be um showing you um the first case is one we did uh in like i said in the bronx looking at a tactic called kettling that the nypd has used and used in this case uh there's a group called ftp four which had organized a rally in the mont haven neighborhood of the bronx and it was really not just the tactic of kettling but the confluence of this tactic with de blasio's curfew and the urban environment that needed to be understood synthetically right and this is kind of all coming together in this analysis i'm going to play a short video um and i want to just point out at the beginning that you know ftp 4 organized this rally uh de blasio's rhetoric in the months and weeks the weeks and months that followed was that the protesters were violent the tactic was uh justified and really this work is about kind of unpacking those claims and holding people holding uh the mayor and and the police accountable for for those statements i want to play a quick video this is a project we did with human rights watch as the marchers headed down willis avenue more than 50 police officers locked the stream yo we gonna go around the march redirected down 136th street and in the final minutes just before the eight o'clock curfew instead of allowing or even directing the marchers to disperse the nypd diverted its bicycle officers to block the marchers just as they reached the intersection of 136th and brook avenue [Music] get down and from behind the march a line of officers blocked the protesters from turning back it's a tactic called cuddly so you know it's the basic idea is that we have [Music] the same event captured from multiple perspectives and with 5g that's only going to you know increase increasingly be so and and it's really about understanding what this kettle looks like from above what it looks like on the ground and and sort of looking at it and measuring it in relation to what the official narrative actually is right um and i want to just comp and that was an advocacy case that was that was that was not meant for court it's meant to put pressure on on um accountability right and redress this is a case that was for court and a kind of gold standard for us in terms of capture here you have a photogrametric scan of an entire urban area uh from a drone that's combined with a series of laser scans on the ground so we can get really high fidelity both on vertical surfaces and everything you might want from a photogrammetric perspective but it provides um very very high resolution information um and allows us to kind of take apart the digital model and i'll try and move through this quite quickly but isolating features of interest and in a legal case it becomes very myopic very quickly and somewhat absurdly but we may be interested in very one very small and very specific question depending on who the plaintiff is um in this case the plaintiff was a group called don't shoot portland um these folks um who organized the protest and we were working with them and we were working with the legal team they filed the case um you know arguing that the use of force by the portland police bureau was um was excessive um and and and used one event to uh a series of events in one evening to try and demonstrate that point um and i'll show very quickly i think i'm gonna do it without sound uh we had to build a tool that was basically not a video not an advocacy video a linear video like you just saw that was used um for the human rights watch case but something you could click on the lawyers could kind of advance to different places the defense could query the prosecution could query et cetera so much more dynamic and interactive format for court but this is the end product a kind of combination of that high fidelity laser scan with video footage which documents very specific incidents and really the question in the legal context is what was the 30 seconds or 60 seconds that led to this sort of use of force or moment of violence um that is what has to get unpacked in a very granular very myopic way and we do this through nine events um in this case some of which also kind of leverage the uh the digital model and the high fidelity nature of it to do kind of a quantitative analysis you know in this case you're getting direction from the portland police bureau move east you can see that this protester is moving east he's complying uh at the same time a flashbang grenade is thrown directly at him and he's injured right so how do we get to the very specific nature of the question when you're talking about crowds and again that the kind of conceit of complexity becomes uh something to hide behind um and i just want to close in terms of the impact of this work i think both of these cases we feel like we moved the needle a little bit in the in the portland case the judge held the portland police uh in contempt of don't shoot portland's uh uh case and and in new york case there was you know a lot of pressure on de blasio and others to kind of to answer for what they've been saying what happened to may and june and i look back with remorse i wish i had done better i want everyone to understand that and i'm sorry i didn't do better and i've learned a lot of valuable lessons and i want our police department to do better um and so just a very very quick reflection on kind of like what's next and i'll end here i thought it was really interesting to see the very kind of mainstreaming of these visual investigations tools on the floor um you know combining diagrams combining videos i think this is only going to increasingly be the case um of course this is a double-edged sword um you know we're doing this because we're passionate about combating abuses of power but it's very easy to see how these same technologies get co-opted and reversed you know for let's say quote-unquote uh public safety um which is something that concerns us and we're getting a lot of somewhat unwanted attention from people who are in that sector and i'll i'll stop there thanks so much brad um super interesting and impactful work um and as i'm you know like looking at all the presentations something that came to mind as i was thinking about patrick gettys i know this sounds like a strange old reference but he was a scottish biologist sociologist geographer philanthropist an urban planner but what i always found interesting about patrick getty's is he really struggled to see the world and find ways to see he tried maps um and have different formats collage and then uh he created his camera obscura which is uh this outlook tower uh to look at the regional landscape and it what i i felt in all of our presentations we were really how do we use this new technology really to see the world in different ways um and kind of really make connections both um good and bad for that another thing that came to mind as i was looking and thinking about all the presentations is um really the promise and perils of new technology and you know whenever we have a new technology and i think sheila's talk really helped uh to amplify that right like even during time when we were getting electricity's kind of debate between the promise and peril of new technology um then how do like i had this question how do we encourage innovation while the same time consider the potential perils for seeing and understanding the worlds we live so you know i felt like like all of us have kind of this one side where um you know there's a potential benefit um and another where we're kind of highlighting the apparel so i kind of pass it to a very big broad hard question but how do we how do we encourage innovation but at the same time consider the potential for it to do harm i see nicholas is unmuted himself oh i i was i'm hunting myself for for a shared discussion but i'd say quickly i mean one of the fundamental rules of technology is unintended consequences so to some extent i think the the mechanisms we need to develop are are adaptive because of all the things that 3d 5g and technology are going to do i can almost guarantee you with with infinite respect for all of the wisdom of my co-panelists that the most remarkable effect is one that we inherently can't understand tonight because it depends on uh on questions of scale and scaling and implementation that we just inherently can't see and it's in the nature of complex systems like cities that once they get to a certain scale and certain size they behave in fundamental ways that we fundamentally don't anticipate so i i don't it's not that's precisely not an answer to your question sarah but i i think your question i mean i think it's the age-old question as you point out that we're you know as we're always thinking about and you know like technology is a double-sided sword but as most things in the world right it's a medium that can convey messages good and bad um what you point out um it's a great question i i certainly don't have an answer for it but um i think that in in a place like mit and certainly in educational um institutions i think the question of the the sort of um incorporation of innovation and the ecology of innovation is really a challenge and um can we uh are there mechanisms whether those be governmental or policy um to bring corporations to a sense of a common good maybe even a common planetary good rather than a corporate good because i think that is a defining characteristic the increasing privatization of um infrastructure service providers certainly in the internet and social media space that do operate as quasi states as we saw with the banning on twitter and so forth um yeah i'll just add that i think in in our practice um it's hard for those things to kind of live in exactly the same place like you know with human rights work often it's you don't really have a chance to get it wrong right there's the kind of do no harm model and uh that can also produce a chilling effect on how we think about innovation and and some of the new tools and so i think what's been most productive for us is to have very robust very active conversations between people who are experts in apparel side of things and people who are pushing hard on some of the ways in which it helps with accountability and redress and the positive side of things and for those conversations to be robust and continuous forums like this i think are very productive but i personally find it nearly impossible to reconcile those two things within one house within one brain they often live separately and in discourse i just want to maybe to remark on the role of the user for uh for essentially accepting of rejecting innovations right if you think about consumer products at the end of the day especially on the user to decide which technology is good or bad right and um you know if i like the iphone i'll use the iphone if i like another phone i will i will just reward a different innovation instead of the iphone i think the problem that you are facing here is that the user is no longer provided with the information which is needed to decide right i'm giving my data and giving up my data because i'm using social media but i'm not aware of that entirely right so i think the first step like you know making sure that we can empower the user and the user in in turn can just prioritize the right type of innovation which is beneficial for society it's just to provide information to make sure that the user is informed that's the bit i wanted to add i guess i i would dive in there because it relates actually to an interesting internal mit question which is this one we're having at the moment around the role of conversations about design across the institute and so uh when it comes to um a lot of uh and here i'm i'm relating to my context in the bay area um particularly in the discourses around design in in the in silicon valley uh so much faith is placed in the user the user is at the center of user-centric design and the uh uh and the user uh uh and and and so much time and energy is spent on on the micro scale of features that that smooth and and uh uh improve our our lived experience but there's a difference between a user user uh centric design and civic centric design i think this you know relates to sarah's agenda and sarah's work and one of the i can say is someone who
2021-05-10