Conveniently connected: tracing the presence of the Cloud in Uber’s services through patent analysis
Article by Piotr Pawel Skora.
Abstract: The idea of convenience, constantly wielded by platforms like Uber as an attractive feature of their services, is built on several assumptions. One of them is the idea of the “cloud” as an unproblematic, ethereal, and comfortable connection between users, providers, data analysis systems, and machine learning. While the reality of the cloud includes precarious work, fragile materials, and fragile devices to be maintained, its purified and abstract conceptualization remains in its imagination. The pervasiveness of this image can be observed in the central role it plays in the configuration of online platforms, which rely on the cloud for their services, and reinforce their appeal as convenient and easy-to-access intermediaries. The present study draws attention to this role of the cloud and turns to patent analysis to make explicit its presence. While patents present themselves through technical and seemingly objective language in their descriptions, it is enlightening to find traces of the ideologies behind them through critical analysis. Thus, this paper employs patent analysis, targeting specifically the patents registered by Uber. It is shown how the apparently neutral and objective depiction of an invention within a patent remains charged with the meanings that the cloud carries within its frames.
Keywords: Platform economy, Cloud, Patent Analysis, Uber
Online on-demand platforms are based on the idea of convenience. The ease of using an application that acts as an intermediary between a seeker of a service and its provider lies at the root of their appeal. Further, the apparent immediacy and omnipresence of these intermediaries reinforce the idea of convenience even more. These features appear to be closely related to one of the crucial elements that enable the operations of online platforms: the cloud. Beyond pure utilization, is there any other relation between online platforms and the cloud? The present study attempts to answer this question by examining the treatment that the cloud receives from these platforms. Specifically, it employs patent analysis in search of the ideological traces behind the usage of the cloud. The company Uber was chosen for this work due to its great number of patents and the specificity of its business model, being a lean platform (Srnicek, 2017).
Gig economy and online platforms
The importance and impact of online platforms that connect users with providers of services like transportation or food have been broadly discussed at this point. For instance, Barratt et al. have recently argued that these forms of intermediation are, rather than passive, playing a “critical role in the creation and regulation of both product and labor markets” (1655). Furthermore, gig economy altogether shapes and delimits the potential agency and behavior of the workers (Barratt et al., 2020) and even users (Törnberg & Uitermark, 2020) that participate in it. Such limitations, along with the new business models created by gig economy, lead to several different forms of precarity of labor (Moisander et al., 2018).
But what are online platforms? Following Srnicek, online platforms are “digital infrastructures that enable two or more groups to interact” (2017, p. 43) and act as intermediaries between individuals that offer or seek a certain service or good. While heavily relying on the idea of freedom and convenience for the user and the provider of a service, online platforms configure new forms of control at the same time, a paradox previously discussed by Möhlmann and Zalmanson (2017). However, even if at the core of these platforms lies convenient intermediation, Srnicek also draws attention to the various business models these platforms follow. Specifically, this study focuses on the category of “lean platform,” defined by outsourcing all the elements of a business and configuring themselves as completely “virtual” (Srnicek, 2017, p. 76). Elaborating on the statement made by Srnicek that lean platforms rely heavily on the cloud for their activities (2017, p. 83), this study argues the cloud remains at the center of their conceptualization.
Frequently, algorithms are claimed to be “impartial and value-free,” while they prioritize the needs of the companies that use them above the condition of their workers (Sun, 2019). Online platforms’ ethereal character enhances this impartiality and neutrality since they rely on the (apparently) immaterial entity that the cloud is. Hu (2015) initially defines “the cloud” as “a system of networks that pools computing power” (IX). However, the author of A Prehistory of the Cloud points out how a complex of electronic resources has turned into something more understandable in our daily language, “a single, virtual, object,” which is “the cloud” (Hu, 2015, p. X). As a result, the cloud’s general understanding is more related to its cultural dimension rather than its technicality. The meanings attached to this concept, namely convenience, limitless space, immediate access, etc., have been generated within this social imagination of the idea instead of being based on an existing material infrastructure that would de facto bear those features. In fact, the cloud shows itself as fragile and very limited to anybody that approaches it closely (Hu, 2015).
How close (or rather far) is the popular imagining of the cloud as a concept to its real materiality has deeper and more problematic consequences. It is crucial to notice that this disjunctive produces a displacement, especially of the labor and material effects of the infrastructure that sustains the cloud itself. The workers who maintain and clean the servers, or the moderators of websites and forums, are just some examples of the labor that remains hidden behind the convenience of the cloud (Hu, 2015, p. XII). Furthermore, as John Peters reminds us, clouds are also not maintained with thin, infinite air, but instead consume vast amounts of electricity and “produce an enormous amount of heat” while doing it (Peters, 2015, p. 332), due to the data that they contain and process (Srnicek, 2017). Acknowledging that “the cloud’s infrastructure was responsible for 2 percent of the world’s greenhouse gas emissions in 2008” (Hu, 2015, p. XXV) means to acknowledge how far the cloud is from being an ethereal or infinite resource.
The World Intellectual Property Organization defines patents as “an exclusive right granted for an invention” (WIPO, n.d.) and, as such, must be publicly disclosed in an application, using technical language (WIPO, n.d.). Thus, patents are commonly imagined bearing a set of associated meanings, objectivity, and technicality, among others, and public patent regulating institutions, such as the WIPO, reinforce such a view. Nonetheless, a closer look at these documents may reveal some of the ideological traces at the base of the invention being patented. Following Shapiro, this article also defends that “patents reveal key assumptions and design choices baked into technology – or at least formalized in the recipe – that are often hidden from plain view” (Shapiro, 2020, p. 768). As a result, patent analysis has the potential of making such ideological choices explicit and available for public debate, taking them away from the aura of technicality and neutrality that impregnates published patents.
The usage of patent analysis in social sciences is rare but certainly not unique. Delfanti and Frey critically approached the patents registered by Amazon in order to show how the human workforce necessary in a warehouse is not about to disappear anytime soon. On the contrary, a thorough patent analysis revealed the intention of Amazon to extend and refine its mechanisms of control over human labor (Delfanti & Frey, 2020). Similarly, Hlongwa has criticized the massive data retrieval that online platforms perform while offering their services, building his argument on the evidence found during the analysis of patents (Hlongwa, 2020). Such an analysis also allowed to show the rapid growth in the development of technologies related to data extraction, including machine learning or database maintenance (Hlongwa, 2020). Finally, it is worth mentioning the article by Shapiro, who draws attention to the futuristic potential of patents in relation to the concept of “smart city” (2020). By conducting patent analysis, combined with critical theory, the author makes explicit the usage of patents for claiming the right to shape urban spaces in the future by applicant companies (Shapiro, 2020).
Uber and Patents
Following the categorization made by Srnicek, Uber is a lean online platform (Srnicek, 2017), meaning it relies on the cloud for all the intermediation services it offers. These services include but are not limited to providing an online intermediary system between seekers and providers of transportation and food delivery. The company has constantly been acquiring (Phelps, 2018) and applying for patents since its foundation, making it the perfect subject for patent analysis. While Uber has been extensively studied in the past, due to the several controversies and unrest surrounding its business model (Möhlmann & Zalmanson, 2017), the patents registered under its name remain out of the frames of these works. The present work attempts an analysis from a micro-perspective of several patents owned by the company to address this gap. The goal is to identify if the presence of the cloud, and the meanings culturally associated with it, have a central role in a medium that is supposed to be free of ideology and cultural assumptions.
Several layers of information comprise patents. Initially, most of them provide an abstract, a title, and a detailed description of the described invention. It is worth noticing that patents are also rich in other kinds of information, as “important dates, inventors, assignees, classification codes and other references cited” (Hlongwa, 2020, p. 49). However, the first part of the analysis mainly focuses on the written descriptions of the inventions, while the second concerns the graphic representations employed in the documents.
Figure 1. The diagram describes the invention registered in the first analyzed patent from the most general perspective. (Yao & Cai, 2020, Figure 1., p. 3).
The first patent is a system, which “uses machine models to estimate trip durations or distance” (Yao & Cai, 2020, p. 1). By employing machine learning, the registered system claims to analyze the latest information received from real-time databases and create a real-time model of a trip. Also, the system may draw information from a database that contains the history of previous travels. Finally, it is claimed that the invented system combines both sources of information to accurately estimate the duration and distance of a trip (Yao & Cai, 2020). The cloud here, similarly as in the rest of the analyzed cases, is represented with the word “network,” and it is present mainly in the general views of the invention, as an intermediary between the “user” and the “system” (Yao & Cai, 2020, p. 3). However, the depiction of the patent focuses on the “system” itself, leaving the “network” with a short and vague definition:
the network 120, which may comprise any combination of local area and wide area networks employing wired or wireless communication links. In some embodiments, all or some of the communication on the network 120 may be encrypted (Yao & Cai, 2020, p. 12).
Nonetheless, the possibilities of the defined “network,” however vaguely described, are similar to the concept of a limitless and omnipresent cloud. Further, the graphic depiction of the network certainly points to the cloud as the concept being discussed in the patent, as shown in Figure 1.
Figure 2. The graphic depiction of the second analyzed patent also draws a symbolic cloud while referring to the “network” (Berger et al., 2019, p. 2).
The second analyzed patent refers to a system for predicting “the value a provider might receive for providing services within the geographic region during the set of time periods” (Berger et al., 2019, p. 1). In summary, the invention is claimed to calculate the value of a service through real-time and historical data analysis and show the results through a range of graphical user interfaces (or GUIs) afterward. In this case, the “network” is defined more accurately, including different networks like Ethernet or 802.11 and network protocols like MPLS and HTTP, among others (Berger et al., 2019, p. 13). However, this definition de facto mentions most of the possible protocols and networks that make the cloud possible in the first place, and thus, configure it. In the same way as the first patent, this “network” lies in the middle of the interaction between the invented system and its users, enabling the necessary communication between them. The graphic depiction of the “network” is, again, a clear allusion to the cloud, as shown by Figure 2.
Figure 3. The “network” is an organic part of the interaction between the “user’s” device and the patented system (Rahematpura et al., 2020, p. 7).
Finally, the third patent chosen for this study focuses on a system that “operates to receive service requests from multiple requesters” (Rahematpura et al., 2020, p. 1). The invention is claimed to select, suggest, and distribute transport providers for two or more requesters simultaneously, considering parameters like the distance between the current location of the providers and requesters (Rahematpura et al., 2020, p. 1). This document does not define the “network” as the previous patents did, but it still includes the symbolic cloud to depict the concept, as shown in Figure 3.
The analysis of the selected patents shows how Uber holds the cloud, and its features of unlimited and immediate access to a network, at the center of the services it offers as an online platform. Furthermore, it shows how this role is represented within the frames of patents, embedded in technical writing and objective depictions of inventions. While the limited scope of this study does not allow for further generalizations, it seems imperative to keep deepening our understanding related to the usage of the cloud as a metaphor by online platforms to reinforce their narratives of convenience. This imperative comes from the acknowledgment made by Hu that the cloud may be thought of as limitless and ethereal, being nonetheless limited in its resources and material in its effect on human labor and the natural environment (Hu, 2015).
Barratt, T., Goods, C., & Veen, A. (2020). ‘I’m my own boss…’: Active intermediation and ‘entrepreneurial’ worker agency in the Australian gig-economy. Environment and Planning A: Economy and Space, 52(8), 1643–1661. https://doi.org/10.1177/0308518X20914346
Berger, A., Williams, R., Ha, E., Hu, M., Valerio, C., Kim, N., & Yi, A.J. (2019). Predictive Session Analyzer For a Network System (U.S. Patent No. 2019/0080340 A1.). U.S. Patent and Trademark Office. https://patents.google.com/patent/US20190080340A1/en?oq=U.S.+Patent+No.+2019%2f0080340+A1
Delfanti, A., & Frey, B. (2020). Humanly Extended Automation or the Future of Work Seen through Amazon Patents. Science, Technology, & Human Values, 46(3), 655–682. https://doi.org/10.1177/0162243920943665
Hlongwa, L. (2020). The city as an algorithmic formation: Insights from patent data. Work Organisation, Labour & Globalisation, 14(1), 47–66. https://doi.org/10.13169/workorgalaboglob.14.1.0047
Hu, T.-H. (2015). A Prehistory of the Cloud. MIT Press.
Möhlmann, M., & Zalmanson, L. (2017, December 10). Hands on the wheel: Navigating algorithmic management and Uber drivers’ autonomy.
Phelps, M. (2018, April 24). How Uber Built Up Its Patent Strength Quickly. ChiefExecutive.Net. https://chiefexecutive.net/how-uber-built-up-its-patent-strength-quickly/
Rahematpura, M., Nickels, J.M., O’Mahony, E., Guo, D., Sundaresan, V., Seeman, L., & Vosoughpour, M. (2020). Network Computer System To Arrange Pooled Transport Services (U.S. Patent No. 2020/0292335 A1). https://patents.google.com/patent/US20200292335A1/en?oq=U.S.+Patent+No.+2020%2f0292335+A1
Shapiro, A. (2020). ‘Embodiments of the invention’: Patents and urban diagrammatics in the smart city. Convergence: The International Journal of Research into New Media Technologies, 26(4), 751–774. https://doi.org/10.1177/1354856520941801
Srnicek, N. (2017). Platform capitalism.
WIPO. (n.d.). Patents. What is a Patent? Retrieved December 19, 2021, from https://www.wipo.int/patents/en/index.html
Yao, S. & Cai, X. (2017) Generation of Trip Estimates Using Real-Time Data And Historical Data (U.S. Patent No. 10,816,351 B1). https://patents.google.com/patent/US10816351B1/en?oq=U.S.+Patent+No.+10%2c816%2c351+B1