The material consequences of “chipification”: The case of software-embedded cars

Abstract

Today's modern car is an assemblage of mechanical and digital components, of metal panels that comprise its structure and silicon chips that run its functions. Communication and information studies scholars have interrogated the problematic aspects of the programs that run those functions, revealing serious issues surrounding privacy and security, worker surveillance, and racial, gendered, and class-based bias. This article contributes to that work by taking a step back and asking about the issues inherent not in the software running on these chips, but on the microchips themselves. Using the lens of “chipification”—the process by which a device is rendered capable of reading and processing data through the embedding of microchips—this article explores how the integration of microchips also involves processes that are often borne from, replicate, or create troubling power dynamics of their own. It takes light-duty passenger vehicles as a case study into how chipification is radically reshaping such processes as resource allocation, labor flows, and cultural practices around car manufacture, use, repair, and modification. By naming the process of chipification, this article allows researchers to identify and analyze the ways that integrating data processing capabilities into everyday devices is not a frictionless practice, but rather one that implicates a variety of power dynamics within these massive industries.

A major national trade show for the performance racing industry might seem like an unlikely place to see cutting-edge computing technologies. One look around the showroom floor and you’ll see booths for aftermarket vendors of every conceivable part of a car: wheels, seat covers, carburetors, gas pedals, spark plugs, etc. Here and there sit huge engine blocks, entire chasses, and complete body frames. Keep looking, and past the shining metal and sheets of plastic, there are stalls displaying nothing but some circuit boards, or small black boxes sprouting dozens of wires, or digital screens displaying data mapped out in graphs and charts. Take a moment to listen, and wandering among these stalls you’ll hear product salesmen, amateur racers, and independent developers discussing the latest software updates and bug patches. All of these—from the metal frames to the silicon chips to the people who know how to build them and make them run together—are necessary, if often overlooked, parts of running a modern car. In reality, to be at the cutting edge of performance racing, or even the forefront of your everyday commuter car technology now requires deep programming knowledge.

The contemporary consumer automobile contains systems that run the gamut of the different modes of data production and use offered by the integration of computer chips. Some function very simply, taking sensor readings produced in one part of the car to modify the activities of another part; the data produced is not stored, but used for calculations on the fly. Others are more complex, capturing and storing data for system diagnostics, assessment, and maintenance. In the most sophisticated models, data is not only stored, it is transmitted, for the purposes of navigation, entertainment, or driver surveillance.

What all these forms of data production and use have in common, at the root, is that they run on computer chips—computer chips that were manufactured, acquired, installed, maintained, and eventually disposed of—but whose materiality is often taken for granted by the ways that marketing, cultural trends, legal discourse, and even academia have long preoccupied themselves with the data they produce. There is a rich literature that engages with the material entanglements of data production and processing (to name some starting points: Cubitt, 2017; Hogan, 2015; Starosielski, 2015). What this article adds is an attention to the moments where the incorporation of data-processing capabilities represents a shift in artifact design, manufacturing, use, and disposal—it’s about the devices that existed in analog configurations before they became digital. It urges our fields to ask: what is unique about the moments when digitization represents a material shift from prior configurations? How does this shift impact different moments in the life of these devices?

To that end, the article introduces the concept of “chipification,” the process by which a device is rendered capable of reading and processing data through the embedding of microchips. The process of chipification includes several steps, most notably the installation of computer chips, the design and uploading of a computer program, and the contexts of use wherein the software is actually run. Chipification occurs any time a device gets a microchip installed in it, either as part of its initial design, or as a later update, but as a framework is best applied to objects that once existed without microchips but which, given recent trends toward digitization and datafication, are being embedded with microchips to allow for digital functions and services—as we are currently seeing across home devices (particularly kitchen appliances), children's toys, and automotive technologies. Thus, chipification is a physical, material process involving hardware like semiconductors, physical labor like mining, manufacturing, and programming, and material cultural practices like contexts of use, and repair, maintenance, and modification activities.

This article guides the reader through some of those threads, to highlight three moments where chipification is having significant impacts on a major global industry. It begins with a literature review of communication and information studies scholarship that interrogates the creation, analysis, and application of data, to show how the questions that this literature raises about power dynamics can be productively applied as well to the material dimensions of data technologies, using scholarship about automotive technologies and infrastructures as an example of such an application. The article continues with a brief history of how automotive technology evolved toward supporting datafication, and then proceeds with three brief case studies of sectors of the automotive industry that would benefit from a critical attention to the changes rendered by chipification. Finally, the article concludes with a discussion of how research into chipification would not only provide new insights into these areas, but would also benefit our own disciplines by building out new capabilities and providing opportunities for important interdisciplinary collaborations.

The chipified car in communication and information studies literature

The modern chipified car runs software for multiple purposes: to calculate the most efficient air-fuel ratios for each engine cylinder, to monitor tailpipe emissions, to provide entertainment and navigation, to report systems functions to the manufacturer, and so on. Fundamentally, the point of these software is to collect, analyze, and relay data—data that is always already laden with assumptions and biases (Gitelman, 2013). The transformation of social phenomena into data has been termed “datafication,” (Mayer-Schönberger and Cukier, 2013: 78). Communication, media, and information studies scholars have revealed the various ways that the datafication of different facets of social life has resulted in injustice due to the fundamentally biased nature of data collection, analysis, and application, particularly along racial, gendered, and class-based lines (Benjamin, 2019; Noble, 2018). Scholars have also revealed how these systems are often premised on profit models that encourage, if not outright require, privacy-invading designs (DeNardis, 2020).

As these information-processing capabilities have been applied beyond personal computers and Internet services, and integrated into other mundane domestic and consumer devices, researchers are finding that these systems create or exacerbate power imbalances between device users and device designers/manufacturers. This is particularly true where those devices are necessary for certain kinds of labor; the software embedded in those devices then becomes an avenue for employee surveillance. In their analysis of recent patent filings, media studies scholars Delfanti and Frey (2021) have noted that, while not yet a functional reality, international megacorporation Amazon is clearly making moves towards embodied worker surveillance via wearables like wristbands. Such surveillance has long been a functional reality within various automotive-related jobs. It has been part of the very business model of such rideshare companies as Uber and Lyft (Rosenblat, 2019), for example. Information studies scholar Karen Levy's work (2014, 2015) traces a lineage of electronic worker monitoring that emerged even earlier than that, starting in the 1980s, in the use of electronic on-board recorders to track the activities of long-haul truckers in the United States.

However, this work on data processing technologies often elides the materiality of the technologies themselves. Writing specifically about the waste created by digital technologies, media studies scholar Jonathan Sterne noted, “[e]mphasis on virtuality, the ethereal, ideational, immaterial, and experiential dimensions of new media leads many writers to accept the myriad strategies that states, institutions, and individuals use to move computer trash into the backspaces of modern life” (Sterne, 2007: 16). This elision is an issue that interdisciplinary scholars in media studies and science and technology studies have warned about for decades, particularly regarding issues of resource extraction, supply chain logistics, and e-waste. Such work has revealed the extensive ecological impact of media technologies, including the incredibly toxicity produced by the mining, extraction, and refinement required by digital components (Bates, 2020; Cubitt, 2017; Parikka, 2015). Most recently, in her book Atlas of AI, artificial intelligence scholar Crawford (2021) documents the impact of rare earth mineral mining on surrounding ecosystems, demand for which is fueled in part by the boom in cloud computing and AI services. Other work has examined the other end of the digital lifespan. Some scholars have focused on the practices and communities of digital device repair, particularly regarding mobile phones, in parts of the world where mobile phones are a financial and professional necessity, but where the parts and schematics may be difficult to access (Rifat et al., 2019). Other scholars have studied how discarded electronic components, which can leech toxic chemicals after their disposal, are often shipped to vast fields in Asia and Africa, where local laborers work in dangerous conditions to extract and re-sell raw material from the circuit boards (Little, 2019).

But how do we instead study those moments where digital artifacts represent a shift within extant device worlds, rather than being sui generis? Most recently, Zane Griffen Talley Cooper's proposed framework of “data peripheries” begins to provide a means of addressing this (2021). In his article proposing the framework, Cooper recognizes that the material entanglements of digital media objects “leverage and depend on a multitude of social, political, and material conditions that proliferate well outside the standard analytical purview” (p. 2) of media and infrastructure studies, particularly those interested in the materiality of information and data.

This article builds off the concept of “datafication” to suggest the concept of “chipification”: if datafication is the process of turning things into data, then chipification is the process of turning things into devices capable of processing data via the integration of microchips. ¹ By identifying and naming these processes behind that integration, chipification can now become a site of study, whereby communication and information scholars can examine the implications of sourcing, installing, maintaining, and repairing the chips that allow devices to run software and process data. While devices that have run software from their origins do go through chipification, what this framework really provides is a tool for interrogating how industries shift from analog to digital: how embedding chips changes fundamental aspects of an industry's material functions, including resource management, labor flows, cultures and contexts of use and disposal, etc. Of course, the labor structures, supply chains, and contexts of use for any digital artifact do not simply emerge from the aether; as much of the above work on extraction shows, the commodity pipelines that electronics come from drew from long-established mines, for example. However, within the fields of communication and information studies, little work has yet been done to understand how the addition of software to devices that previously had none necessitates material change throughout the pipeline, very often in ways that reify existing inequities or bring new inequities to the fore (Maxwell and Miller, 2008).

Focusing on the material impact of software is a natural extension of scholarship on the material impact of cars and our interest in information processing technologies. Thus, to explore this new framework, I apply the concept of chipification to the realm of light-duty vehicles, which covers most consumer vehicles. ² While the car might seem a counter-intuitive artifact of study for the communication or information science scholar, its seeming out-of-placeness is precisely what allows it to reveal the importance of studying chipification from these disciplinary perspectives. Cars have long been understood as influential technologies for shaping urban landscapes, transportation infrastructures, and mobility (Featherstone et al., 2005). Historians point to the car as the main facilitator for the development of suburbs (Vinsel, 2011). On a national scale, in the US, cars have long been symbols of individuality and freedom, seen as the ultimate liberal technology (Vinsel, 2011: 5). On an individual scale, cars have also been seen as powerful icons of consumer identity, nationalism, and space-making, particularly in minority communities (Chappell, 2012; Cross, 2018). The rise of the gig economy, automation, and the Internet of Things has brought particular attention to the software components of cars, particularly as they extend existing concerns about worker surveillance, privacy and security, and bias in automated systems (Rosenblat, 2019).

This body of work indicates how impactful the car has been across a variety of scales and situations. To take that insight a step further means examining the car as an artifact in itself: how have cars been designed and built over the years that lend them to such significance? How have changes in automotive design and manufacturing shifted the car's impact on these various realms? Such work would take the automobile as a starting point, not a given, in its analysis. Levy's (2014) dissertation presents one example of such work, in her examination of the chipification (though she does not use that term) of the tracking of long-haul trucking labor through the introduction of electronic on-board recorders (EOBRs). In it, she traces the transition from paper logs to digital surveillance of trucking, including how the design and in-cab placement of the new digital interfaces of the EOBRs also facilitated new (but by no mean frictionless) relationships between truckers, their employers, and law enforcement. Here, it is not just the data produced by the EOBRs that altered relations, but their very physical qualities.

More to the point, chipification is a process already acknowledged by car culture as a practical, if at this point outdated, way of modifying a car's engine performance. More commonly called “chipping” in automotive communities, this practice is one form of “tuning” a car—that is, modifying a car's engine control unit (ECU, the car's central computing component). Chipping describes the practices of physically replacing the microchips in an ECU, rather than plugging the ECU into an off-board computer, rewriting the programming, and uploading the new program to the existing chip (which practice is known as “flashing”). ³ By borrowing and tweaking this term, I will be building off of car culture's existing and deeply material engagement with these artifacts.

My knowledge about chipping comes from 5 years (2017-ongoing) of ethnographic engagement with mainstream automotive culture in the United States, which included dozens of interviews with car enthusiasts, mechanics, and modifiers and the employees of the small businesses that provide those practitioners with the parts and tools they need to make their repairs and modifications. It also included several visits to the main automotive aftermarket industry trade shows and membership meetings of their representative organizations, as well as regularly reading automotive industry trade magazines and auto culture and technology blogs. This article builds off of conversations I observed in the field where practitioners discussed shifts in their industry that worried them: changes in labor, in their ability to access resources, in the kinds of products and services their customers were demanding. I followed those threads through the industry publications I was reading, and further out into the broader tech press and academic literature.

With cars’ complex legacy thus in mind, I find them to be an excellent case study through which to explore the framework of chipification. By examining how the automotive industry and culture have been affected by the integration of computer chips and computer software into a category of device that was completely mechanical for nearly the first century of its existence, we can begin to explore not just the consequences of digitization broadly speaking, but the specific qualities of the shift from analog to digital, from the mechanical to the chipified.

How the car became chipified

A brief history of how the car came to be chipified already begins to reveal how chipification is both the product of, and itself produces, a complex tangle of policy approaches, design strategies, consumption patterns, and cultural meanings. Although the current trend of automotive chipification can often make the integration of software into automobiles seem like a technological inevitability, in reality, software came to be integrated into light-duty vehicles through a combination of regulatory pressure and technological trends, each of which was motivated by specific material and political realities.

Light-duty automobiles were first mass-produced in the United States starting in the early 1900s, and in spite of some early opposition, were quickly taken up by American consumers and rapidly changed the face of transportation, infrastructure planning, and mechanical engineering in the US. However, by the mid-1940s, it was beginning to become clear to many municipal and state governments in the United States that cars as they were currently designed were leading to unsustainable living conditions, particularly in urban centers where rising car ownership rates were causing smog and other public health concerns. Public outcry led to the passage of a series of air pollution regulation laws in the 1950s and 1960s, in which the federal government (as well as several state governments, most notably California) established a research agenda to investigate automotive emissions, as well as the first emissions standards for motor vehicles. ⁴

Foreseeing a future where automotive emissions were likely to get ever more stringent, auto manufacturers recognized the need to find more efficient engine mechanisms. At the time, most automotive engines used one of two methods for mixing air and fuel for combustion: carburetors and mechanical fuel injection. Both methods only offer limited adjustability, and are unable to take into account conditions in other parts of the engine which may affect the optimal fuel/air mixture. As a result, neither was capable of achieving the kind of fuel efficiency that auto manufacturers needed in order to comply with the new emissions laws. Auto manufacturers needed a better fuel consumption mechanism, one that could respond to more variables more quickly than either carburetors or mechanical fuel injection systems could.

At the same time, microprocessors had become cheaper and hardier, and enterprising automotive manufacturers realized that they were now robust enough to be used in the mass production of automotive engines. The idea of using computers in engine design had been around since at least 1958 when the American aviation company Bendix introduced the Electrojector system, the first electronic fuel injection (EFI) system. The Electrojector was capable of responding to changes in intake manifold pressure, engine speed, air pressure, and temperature, using a “brain box” (electronic control system) to take readings from sensors and use them to modulate the operation of the valves through which fuel was injected (Winkler and Sutton, 1957: 1). However, the Electrojector brain box's electronic system ran on vacuum tube technology, which was particularly unreliable; it was eventually offered in a very limited run of 1958 Chrysler vehicles, but almost all those cars were eventually recalled and had their EFI systems replaced with carburetors.

In 1967, Bosch introduced their own EFI system called the D-Jetronic, which was first available on the 1968 model year Volkswagen 1600s. However, American auto manufacturers were reluctant to adopt this new technology. Some experts attribute this slow uptake to the EFI system's high cost when compared to other fuel injection systems, and differences in American and non-American (specifically, European and Japanese) car trends; EFI provided greater efficiency in smaller engines (Jurgen, 1975), but Americans at the time liked their engines big.

The passage of the Clean Air Act (CAA) in 1970 obligated a major shift in automotive design. The CAA called for a 90 percent reduction in emissions in all new cars starting with 1975. ⁵ Engine manufacturers like Bosch saw this moment as a breakthrough for a broader application of EFI systems (Gorille et al., 1975: 970). In 1975, American automakers began offering the first vehicles with EFI systems as a standard option (Jurgen, 1975; Our Motoring Correspondent, 1969: 34). From there, automakers began to extend computerization to other engine systems. In 1980, General Motors started using a proprietary engine management system, called the Computer Controlled Catalytic Converter (C4), to control not just emissions-related engine functions, but also some additional engine parameters (such as manifold pressure and idle speed control functions), as well as offer self-diagnostic capabilities (Grimm et al., 1980). Other automakers soon followed; by 1983, Ford, Chrysler, Honda, and Nissan had introduced their own proprietary systems (McCord, 2011: 8).

By the mid-1980s, these systems were common enough that regulatory bodies were beginning to take notice. In 1985, the California Air Resources Board (CARB), having realized how these systems could be used to facilitate emissions testing, required that all cars starting in the 1988 model year have some form of driver-accessible on-board computerized engine diagnostics system. ⁶ These first-generation systems came to be known as OBD-I. However, this law did not standardize the codes or physical ports that these systems used. At the time, each manufacturer was using their own connector style and diagnostic code standards (with some manufacturers even using different styles and standards across different models), causing a lot of confusion and extra work for independent shops and owners (McCord, 2011: 9). It also made emissions testing a complicated affair. The federal government addressed this in the sweeping amendments to the Clean Air Act that passed in 1990, which required that all cars starting with the 1994 model year be equipped with standard and uniform diagnostics systems (encompassing ports, connectors, and codes).

Today, the U.S. federal government requires a handful of chipified elements beyond the OBD-II system on all cars manufactured or sold within the US. Notably, this includes an ever-expanding stable of systems that fall under the umbrella of Advanced Driver Assistance Systems (ADAS), a class of automotive systems that provide drivers with varying degrees of computationally supported driving aides. As of the publication of this piece, the only mandatory ADAS systems are anti-lock brakes and electronic stability controls (both required as of September 2011) and rearview video systems (required as of May 2018). In this way, increased chipification is seen as the best, and sometimes only, answer to concerns about vehicle safety and fuel emissions, thus encouraging the expansion of chipified elements in cars.

Understanding that the chipification of cars was driven primarily by concerns about emissions and safety, and whose particular execution was supported by technological trends like digitization and regulatory frameworks like intellectual property (in the pursuit of proprietary models), helps us as communication and information scholars begin to see the cracks where we might begin asking other critical questions. For example, while the concern about automotive emissions was real, does the proposed solution to that problem (integrating software to fine-tune fuel use) present any emissions-related concerns of its own? And while chipifying cars may make them more environmentally friendly emissions-wise, how might the integration of software present other challenges to health and safety, for example in the ways that it alters resource allocation, labor practices, or long-term maintenance and repair possibilities? To what degree does using software to solve these problems introduce problems of its own?

Chipification from three angles

What follows are three sites of study that shed light onto the consequences of the chipification of light-duty vehicles, derived from conversations and topics I frequently came across while conducting my fieldwork: global supply chains, labor, and cultures of use. These case studies are not meant to be exhaustive, either as deep dives of these areas or as a list of topics that need attention. Rather, they are meant to serve as a provocation for further research and an urge to the fields of communication and information studies to re-attend to the logistical, material, and cultural impacts that chipification has had and will continue to have as computing devices and capabilities become more and more ubiquitous.

Global supply chains

The increase in technical complexity that chipification requires also increases demands on the supply chains that fuel the manufacture and dissemination of these products. This impact is felt both in the demand for the raw materials for parts manufacturing and further down the supply chain in the demand for the parts that make up larger components. For the most part, these changes require shifts in existing systems, rather than the creation of all new ones.

While the impact of digital technologies on the demand for raw materials like rare earth metals is somewhat well researched within media studies, as discussed above, less attention has been paid to the downstream impact of chipification on the demand for the parts needed to make a device run software. The most fundamental of these are the semiconductors that make up the computer chips on which the software runs. The proliferation of computers in cars is having a marked impact on the semiconductor market, which has identified the automotive sector as one of the fastest-growing markets for semiconductors, even as other parts of the semiconductor sector's growth plateau due to the saturation of the consumer electronics market (Deloitte, 2019). We can see the ripples of this shift in industry demand in moments of breakdown, most recently during the COVID crisis. In the midst of the COVID pandemic, as manufacturing of all kinds slowed, original equipment manufacturers (OEMs) had to idle shifts and even entire factories due to the shortage of semiconductor chips (Gitlin, 2021). This is partially due to the fact that automotive chipification requires different kinds of chips than most consumer electronics do. While consumer electronics manufacturers focus on the smallest possible chips, automotive manufacturers don’t need to pay for the smallest of the small—cars need chips that are 45–65 nm, as opposed to the single-digit nanometer chips sought by cutting-edge consumer device designers (Culpan, 2021). Due to the automotive industry's relatively small consumption of semiconductors (accounting for only about a 10th of global semiconductor fab output), the sector has struggled to regain momentum, as it was difficult for the OEMs to maneuver their minor market position to recapture output once the factories had started up again (Gitlin, 2021). There is also increasing concern that, because automotive manufacturers represent such a relatively small (and thus less powerful) demand on semiconductor manufacturing, the resulting vacuum will be filled by counterfeit semiconductors whose products could represent huge safety risks (Leprince-Ringuet, 2021).

This push and pull between the automotive industry and the semiconductor manufacturing industry (and the wider digital device manufacturing sector) has consequences both for the manufacturers themselves and for consumers downstream, including not only individual end-users but also ancillary players like dealerships and the used car market. Dealerships, unable to rely on the income from selling new cars due to the manufacturing shortage, have been shifting their resources to parts and services instead (Martinez, 2021). Average used-car prices are up 30% from the beginning of 2020, reaching a record $26,457, with many used cars now more valuable than they were new; because they’re such a hot commodity, used car dealers are having trouble staying stocked (Naughton et al., 2021). These shifts have been pinned on specifically on the shortage of semiconductors, evincing an incredible vulnerability for the entire automotive industry.

While the automotive industry may have trouble jockeying to get their share of a finite semiconductor production pool, this represents a delay for production, not a crisis of access—the concern is when the microchips will become available, not whether. There are other parts that are critical to automotive chipification where the concern is whether components will ever be available at all, and where the fickleness of production demands are having impacts on the ability of car owners to ensure the longevity of their vehicles. In one of my interviews with an aftermarket ECU designer, he mentioned that one of the variables that his company needs to consider when deciding to develop an ECU for a new car model is whether the connector cables for that model are still being manufactured; because most of their clientele buys their hobby cars once they’ve reached the nadir of their resale value, many of the most popular models for hobbyist modifications are several years, if not decades, old. If the connector cables are no longer being manufactured by the OEM, then finding an independent manufacturing facility is usually out of the question, as there simply is not enough aftermarket hobbyist demand to meet the batch quotas that factories require. Instead, the company needs to research whether there are enough of these model cars in the junkyard that they can cannibalize the connectors from those old cars to sell to hobbyists with their new ECUs. While parts availability has always been part of the challenge for automotive enthusiasts, the difference between having to have a mechanical part custom-made, and an electronic part, is vast—actually, for most electronic components, custom-made is not an option.

Chipification also presents unique challenges for the distribution and sale of the entire car itself. Issues with the software itself can add an additional hitch to product roll-outs, a fact of life for most software manufacturers, but an occurrence that still makes headlines for the automotive industry. In mid-2020, Volkswagen had to delay the release of some versions of their first mass-produced electric hatchbacks, the ID 3, due to a delayed software update (O’Kane, 2020); months later, in early March 2021, Volvo experienced a similar situation when it had to hold shipments of its first electric SUV at US ports while they waited for a software update to ship (O’Kane, 2021).

Consequently, we see that chipification is a phenomenon that occurs differently between device categories and that these differences have an impact on manufacturing and fabrication trends. The demand for microchips to embed in light-duty vehicles has had a cascading effect across many different aspects of heavy industry, some of which are directly connected (shortages in cobalt lead to production delays), some of which are more independent (production halts are hard to restart due to greater demand from the consumer electronics industry). These are distinctions that communication and information studies are well equipped to interrogate: to examine how the overall trend towards making all consumer products capable of running software actually has variable roots and effects.

Labor

In addition to affecting the flows of resources and parts, chipification is having an impact both on the labor used to produce the artifacts, and on the labor used to repair, maintain, and modify them. This section focuses on many of the same sites as above—namely, semiconductor and microchip production factories—but rather than discussing the raw materials entering into, or finished products leaving, those sites, here I focus on the impact chipification is having on the labor needed to run those sites.

As with the expansion of any major industrial activity, the spread of chip foundries is having a number of downstream effects on labor and labor flows. Chip production is water-intensive, but some chip-fabrication plants claim to be able to contribute to local water supplies; for example, Intel claims that the water restoration project it is developing alongside its plans for a new foundry in Arizona will actually reach net positive water use once completed (Intel, n.d.). But the placement of these foundries also depends on, and has a significant influence on, the availability of a highly skilled workforce, and can thus influence everything from labor migration patterns to local educational curriculum design. In Arizona, for example, the establishment of chip foundries shifted the priorities of local universities, who developed a reputation for teaching and research in semiconductor design (Salter, 2021).

While the automotive collision and repair industry has always had to evolve to keep up with shifts in automotive technology trends, chipification is presenting a unique challenge, as training to work with highly computerized components requires a different set of skills and continuing education models than the mechanical components of old. In his 2007 dissertation examining how automotive repair education programs respond to changes in automotive technology, Savin (2007) notes that the introduction of microprocessors in the 1970s is what kicked off the rapid speed of change in automotive technology that we see today. However, community and technical colleges (which train the majority of automotive repair technicians) have been struggling to keep up with this pace. The process for adding new technology to the repair curricula often relies on a lengthy certification process by an industry organization coordinating with manufacturers, shop owners, educators, and equipment and parts suppliers (Savin, 2007: 117). Consequently, many automotive repair educators felt like their programs were constantly lagging behind current automotive technology trends, particularly (at the time) those relating to hybrid and computerized systems.

Even with an up-to-date education, repair and modification services working on chipified automotive systems face resource challenges due to the inaccessibility of system schematics or the significant spatial requirements for recalibrating these systems. As I found in my fieldwork at automotive industry trade shows, a topic of frequent conversation at the education track of these shows in the last few years has been how independent repair shops are struggling to repair cars with ADAS technologies, which include features such as blind-spot warnings, adaptive cruise control, and automatic parking. These systems rely on sensors to work, but these sensors must be precisely calibrated to work properly. This presents a number of challenges to automotive repair and modification specialists. To begin, sensor placement and calibration parameters differ between models and years, and some repair shops struggle to access accurate schematics; while some are available through the aftermarket, others are only available from the auto manufacturer, often for a significant fee. Sometimes, the information is nearly impossible to find. What's more, actually calibrating the sensors often requires very particular space requirements: the American Auto Association (AAA) notes, as an example, that Honda requires an open area that is 13 feet wide, 5 feet high, and extends at least 23 feet in front of the car—not all repair shops have this kind of space (AAA Automotive, n.d.). Finally, even when calibration specs are available through the manufacturer, they are only provided for recalibrating the car to factory settings. This means that if a car has been modified in any way—say, for example, by having been lifted, a common modification for pick-up trucks—then the repair shop must determine how to alter the factory settings recalibration procedures for the truck's new height. While some of the aftermarket's industry representative organizations are working together to provide guidelines for lifted trucks, these guidelines still require adjustment depending on the make, model, and year of the truck (Waraniak et al., 2020).

The automotive aftermarket parts industry is also stymied by the opaqueness of chipified automotive systems and the inaccessibility of necessary schematics and source code. The aftermarket industry refers, broadly, to the market of manufacturers, distributors, retailers, and installers of any car parts that did not come directly from the original equipment manufacturer. In order to sell a new product, aftermarket parts makers have to make sure that their product interoperates with the car on both mechanical and programmatic levels. However, as mentioned above, schematics can be hard to come by (either due to price or to availability), and the source code to the software even more so.

(Chipified) car cultures

Cars are often one of the biggest expenses a person expects to make, and understandably most consumers expect to keep their vehicles for a number of years. Due to advances in general automotive technology, the lifespan of light-duty vehicles has consistently risen, with the age of the average vehicle on the road now at a record high of 11.9 years (for comparison, that number was 8.4 years in 1995) (Bureau of Transportation Statistics, n.d.). There is some indication that electric cars are contributing to this increase. Electric cars have fewer parts to maintain or repair, which can contribute to a longer lifespan. However, it may be too early to say conclusively; because the market for hybrids and electric cars only really appeared in the early 2010s (the first Tesla model became available in 2012), we are only now reaching the end of their warranties, which typically last for ten years. What's more, researchers are actually having a difficult time tracking electric vehicle trends across time because advances in battery capacity and vehicle range are happening year-to-year (Taft, 2021).

However, this extension in vehicle lifespans has so far largely been attributed to advances in battery technologies, rather than to any advantage offered by chipification. In fact, to the extent that other chipified devices are canaries in the coal mine, there are some indications that chipification may actually pose a challenge to extending vehicle lifespans in a financially accessible way. As manufacturers invest in frequent updates in software packages and systems, without leaving a support system for older versions of that software, consumers of older devices may be left without access to important functions. In the networked home market, this has happened as smaller companies are acquired by larger ones: in 2014, Google's home automation branch, Nest, merged with a competitor, Revolv, and only two years later announced that it would stop supporting the service for Revolv hubs, leaving consumers with bricked devices (Price, 2016). While the automotive market is less likely to see these kinds of acquisitions, vehicle owners are still experiencing service stoppages as OEMs pivot to newer systems. In 2018, for example, some Ford vehicle owners were informed that Ford was no longer supporting their Vehicle Health Report, which supplied drivers with information about various vehicle components. In their announcement, for stated that “While this technology was useful when the feature launched, it is no longer the best way to deliver vehicle health information to Owners”; apparently, the better way to receive this information was to require owners to visit their dealers (Cooley, 2018; Ford.com, n.d.).

For many observers within American car culture, it’s clear that chipified vehicles are here to stay, and for some, this is creating some concerns about the future of car culture as both a hobby and a livelihood. At the town hall events held at some industry, trade shows that I have attended, many in the amateur racing community have expressed a concern that the more sophisticated in-vehicle computers get, the more intimidated young people are at the idea of tweaking them, thus eroding general interest in car modification and, consequently, in engaging with racing as a hobby. It is also impacting professional car cultures, particularly in the repair and collision sectors. This is because learning to work with chipified cars requires different skill sets, and investments of time, then working with more mechanical vehicles, and these can often seem like barriers to new entrants, particularly those coming from less privileged backgrounds.

The shift to chipification is also facilitating a concerning shift toward subscription models of car ownership, in which a consumer buys the physical car but has to pay subscription fees for different services on the car, ranging from fuel efficiency boosts to heated seats to autonomous driving systems to adaptive cruise control (BMW, 2020; Torchinsky, 2020). Legal scholars studying communication and information technologies have also sounded the alarm that corporations could use systems like copyright to maintain their control over chipified devices like cars, impeding vehicle owners’ ability to tinker with and learn from their cars (Perzanowski and Schultz, 2016; Samuelson, 2016; Tusikov, 2019).

Communication studies’ close relationship with critical cultural studies provides a plethora of theoretical frameworks that can allow researchers to approach the relationship between the cultural and the material while accounting for power. As critical cultural studies scholar Kelly Gates has urged, “Only by understanding the ways that culture is instantiated in physical spaces and material forms will it be possible to make meaningful changes in how those spaces and forms are built and rebuilt” (Gates, 2013: 245). By offering insights into how the physical changes occurring in cars is impacting the cultures around the production and consumption of cars, communication studies could begin to help craft meaningful interventions into those changes.

Conclusion and discussion

Communication and information science scholars must make sure to reckon with not only the new (and often unequal) flows of data and information that chipified devices make possible, but also the flows of labor, resources, and material cultural practices. These flows have already started to radically reshape many major global industries.

These issues are all the more urgent as the world careens into the chipified era. In the automotive sector, it seems chipification is a bell that cannot be unrung. From some angles, this may seem like a good thing. The chipification of the internal combustion engine allowed for the kind of data capture and processing that granted ICE engineers the ability to fine-tune fuel ratios and thus lower engine emissions, with average vehicle emission rates per vehicle, for every kind of greenhouse gas, decreasing significantly since the early 2000s (U.S. Environmental Protection Agency, Office of Transportation and Air Quality, 2021). The added ability to store and report that data also increased the ability of state governments to enforce automotive emissions standards on individual vehicles. ⁷ At the automotive trade shows I attended, presentations on ADAS technologies touted their role in increasing driver and passenger safety and lowering crash deaths (Waraniak, 2019). ⁸ If light-duty vehicles had not been chipified, could we have achieved the kinds of emissions and safety levels that would have allowed for the continued viability of mass individual car ownership as the main form of transportation in the United States?

A more critical approach to chipification—one that looks beyond the benefits provided by the data these chips produce—might ask instead whether we want to maintain that status quo in the first place, and if so, at what cost? Or, put specifically in the case of automotive technology, are the added stresses to supply chains, labor, and contexts of use worth the gains made in fuel efficiency and passenger safety? To what extent does the push towards safer and more efficient cars through chipification just further entrench cars as the dominant mode of transportation in the first place? How does the push to achieve safety and efficiency goals through chipification obfuscate the ways that integrating microchips also serves to increase corporate surveillance through and control over our cars? Is it better to continue developing digital tech that makes cars safer and more efficient, or instead work toward making cars rarer and less necessary (through investments in public infrastructures, rather than in digital automotive tech)? The concept of chipification allows researchers to identify and analyze the ways that simply throwing processing capabilities at a problem is not, in fact, a frictionless practice (Goulden, 2020).

This is a prime site for communication and information studies to engage and provide critical insights. Practically speaking, it is an area of increasing importance to our students, particularly in information studies. As more and more devices run software, and more and more device manufacturers are pivoting to developing that software in-house, graduates of our programs are likely to be looking for jobs not just at the social media giants or data-processing megacorporations, but also within smaller teams in unexpected places like Toyota, Volkswagen, and Bosch (an automotive parts company), all of whom have announced initiatives to build in-house software operations teams within the next 5 years (Greimel, 2021). Teaching our students to be aware of the consequences of integrating software into a device that has unique contexts of design, manufacture, and use will not only better situate them to design these systems, but to do so critically and inclusively.

The topic of chipification also presents an incredible opportunity for communication and information studies to continue advancing our interdisciplinary ties, making a case for how the work done in our fields is critical for expanding upon or mobilizing the work done in others. For example, communication and information studies could build on and with the e-waste scholarship coming from STS to incorporate e-waste considerations into system design calculations (i.e. is it worth the eventual e-waste to produce a chipified product?). There are also rich possible collaborations with such fields as public policy, engineering, and environmental science, as communication and information science, scholars can provide more insights into the everyday lives of these systems, infrastructures, and devices for use in developing more equitable regulation, design standards, and even pedagogical goals. Moreover, these interdisciplinary ties can, and indeed must be, cross-national. As a scholar trained in the United States, my citations are unfortunately largely drawn from scholarship published in American research institutions; however, there is a great deal of compelling work occurring in Australia and Europe on the material consequences of the Internet of Things, particularly in domestic settings (Bates et al., 2014, as a starting point).

Chipification takes up Sterne's call to attend to the material consequences of the most ephemeral-seeming communication technologies of our time. It takes as a moral imperative that communication and information studies begin to interrogate these devices and their attendant processes and practices. By focusing on the impacts of these technologies, these fields’ current approach to emerging chipified technologies often takes for granted, and thus runs the danger of implicitly reifying, the many ways that these technologies efface their material impacts (Sterne, 2007). Making chipification the question, rather than the given, allows us to turn those assumptions on their head, so rather than asking, for example, “How might we build better/safer/more ethical data applications and flows?” we might ask “Should we be building the capacity for data flows into this device in the first place? What are the material costs of doing so, before the first bit of data even gets produced?”