3While there has been some literature on projects from the JNNURM and NeGP (Chatterji 2018; Kundu 2014), these works evaluate the progress and completion of projects across states under the larger missions and not the implementation and impact of technologies or digitalization practices at the city level. We only see analysis of digitalization policy at the national level (Parkar and Purandare 2023), some city-level, or technology specific studies during the Smart Cities Mission (SCM) (Parkar, Zérah, and Mittal 2023; Praharaj 2021; Prasad, Alizadeh, and Dowling 2022). However, these studies do not refer to the municipal transformations in functions, service delivery, and capacities or the impact of technologies on governance. This gap in understanding how the technological aspect impacts the practice of urban governance is quite clear in the Indian case. For this paper, digitalization is defined as the way in which the transformation of tools, ideas, and practices brought on by technological interventions affects urban governance.

4The literature has pointed out that “technological urban solutionism” tends to replace citizen participation and drive urban planning (León and Rosen 2020; Morozov 2013; Sadowski and Maalsen 2020). While the role of consultants in urban planning and policy is not new (Kennedy et al. 2014; Kennedy and Zérah 2008; Sundaram 2019) and the push for consultants under the Smart Cities Mission has been noted (Khan, Taraporevala, and Zérah 2018; Purandare 2021), the role of consultants has moved away from presenting policy briefs requiring bureaucratic or legislative approval in favor of consultant-driven technologies completely replacing entire governance processes. The digital interventions explored here are clear examples of this transition, whereby both the identification of problems and the tools offered for governance come from existing tech industry trends and not necessarily from democratic or bureaucratic channels.

6I will begin with an introduction to the urban expansion of Bhubaneswar, the administration governing the capital city, and some of the digital initiatives driving transformations in the city and the state. I will go on to elaborate the methodology that guides this work, which disentangles the development and implementation of digital technologies for urban administration by interviewing officers and staff from public agencies as well as technology designers and developers. I will then focus in detail on how the three technology platforms function. I will conclude by drawing on the interviews and evaluate the platforms to present my arguments.

7This paper attempts to establish a method for unraveling the complexity of digitalization in the context of urban governance. It suggests that while motivations for digital projects are grounded in administrative logics, its implementation is entirely beyond administrative control. For researchers aiming to assess the impact of digitalization, it is useful to triangulate these motivations, how the design and implementation of digital infrastructure copes with digital legacies, and the actual user experience of technology platforms at state and citizen level.

The Case of Bhubaneswar

10The three initiatives chosen for this paper include the Bhubaneswar Land Use Intelligence System (BLUIS), the Building Plan Approval System (BPAS), and the Digital Door Numbering System (DDNS). These three governance tools were chosen for the following reasons. First, they specifically target the governance of land use and housing, a topic that has preoccupied the state administration. Second, while they were deployed by consultants and vendors, the platforms were already integrated by the state and municipal agencies—unlike the SCM projects, which still had not been handed over to the municipal agencies and were managed by consultants and vendors at the time of writing. Third, they can be thematically grouped as GIS-ICT tools for urban governance and, as such, have certain common characteristics that will be identified.

Studying Design and Practices of Technology in Governance

11Drawing on Orlikowski (2000), I seek to understand the structuring, modifications, and implementation of “technologies-in-practice.” To do so, I conducted nineteen interviews across institutional hierarchies at municipal and state department offices in Bhubaneswar over three visits between March 2021 and November 2022. The interviews with public officials were carried out at multiple levels. Elite interviews (Dexter 2006) were conducted with one state Housing and Urban Department bureaucrat, two Bhubaneswar Development Authority officials, two senior municipal administrative officers, two municipal engineers, and two officers from the Smart City. Three municipal staffers who handled the platform’s day-to-day operations were also interviewed.

12In addition, it was necessary to understand how technologies are developed, adopted, and implemented—specifically, how platforms and algorithms are developed, what determines database attributes, and how they are used within everyday governance practices. One of the methods Kitchin (2017) identifies for studying algorithmic platforms is interviewing designers and programmers as a way of “uncovering the story behind the production of an algorithm and to interrogate its purpose and assumptions” (p. 24). Interviews were therefore conducted with two vendors, two technology developers and programmers, and two consultants who are responsible for developing platforms and training staff how to use them. One interview was conducted with a subcontractor responsible for importing older municipal databases into a newer platform. I supplemented these interviews with municipal documentation, such as webpages and training videos that visually demonstrate how these platforms and algorithms function. Finally, notes on the observations of citizens accessing municipal offices and staff interactions with citizens are used to highlight how these technologies are experienced.

Spatial Governance through Technology

13I study three spatial technologies specifically targeted toward land use and housing governance. The Bhubaneshwar Land Use Intelligence System (BLUIS) monitors encroachment on public land using satellite data, and the Building Plan Approval System (BPAS) automates the approval of building construction plans, while the Digital Door Numbering System (DDNS) allots a specific digital code to all households in order to streamline and unify the delivery of municipal services. While BLUIS and BPAS are fully functional, only DDNS’s pilot projects have been rolled out in some wards of the city. I will begin with a description of each of these platforms and go on to highlight the challenges and lessons from their use within existing municipal institutions and processes.

15Figure 3

Complaints on BLUIS 2020–2021
Source	Citizen	Officer	Automated	Total
Total	124	0	531	655
Resolved	70	0	350	420

Recorded Instances of Encroachment Complaint by Type on BLUIS (BDA 2020–2021).

17While BLUIS is aimed at tracking illegal constructions, the BSCL suggested the need for faster approval of legal civil construction, which led to the development of an (online) Building Plan Approval System (BPAS) for Bhubaneswar. While BPAS was originally imagined and implemented under the Smart City projects, it was limited to the city territory and was developed as a licensed proprietary product. This first BPAS platform was replaced by a new BPAS for all Urban Local Bodies (ULBs) in Odisha under the statewide platform Sujog in 2022. Regardless of the BPAS model being viewed, it functions in the same way. Architectural files for civil construction are uploaded by landowners or approved architects to an online portal, and an algorithm screens them for violation of building and planning codes. This algorithm encodes existing building plan laws as rules and “reads” the submitted files to look for aberrations. If any conflict is found, the application is either rejected with a note requesting rectification or forwarded to an officer for human scrutiny. The Sujog BPAS also features preapproved templates that citizens can use for construction. The online procedure is supposed to reduce corruption, speed up approval by reducing human oversight, increase transparency through monitoring dashboards, and deliver approval certificates digitally. Figure 7 shows that the system’s algorithms automatically flag nearly half of the pending submissions as “defective,” while the rest are under various processing levels in the escalation matrix. Based on this table sourced from the official dashboard, it is unclear how many of the approved applications were found defective at the time of submission.

18Figure 7

Building Plan Approval System (2020–2021)
Applications Received		1258
Pending	Total Pending	726
	New	25
	Under Processing	129
	Found Defective	348
	Resubmitted	69
	DPBP	19
	Show Cause	72
	Demand Notice Generated	62
	Sanction Fees Paid	02
Closed	Total Closed	532
	Approved	507
	Rejected	25

Process-Wise Escalation and Resolution of Cases under the Older Building Plan Approval System (BSCL, 2021).

19Household addresses are typically labeled through a postal coding system along with local land and revenue registration details. In recent years many cities (such as Hyderabad and New Delhi) have been implementing a Digital Door Numbering System (DDNS). The DDNS project in Bhubaneswar is the third platform analyzed in this paper. The DDNS project is currently being rolled out only in some wards and is not yet completely live or integrated. It is being implemented by a consultancy firm and subcontracted to two vendors, one providing satellite data and the other providing proprietary technology for the location codes.

20Figure 8

21The DDNS system tags all households with a 14-digit alphanumeric code that is geolocated and issued to all premises. This coding allots a specific key that binds each household to their municipal services. A particular household’s municipal billing and taxation is tied to this code and allows the municipal agencies to obtain a household-level picture of service consumption and tax collections. The BMC envisions that DDNS will streamline revenue generation, since the household will be mapped with property tax, commercial licenses, waste collection, and electricity and water meters. Any household misusing services or using commercial licensing will be identified through this. Households that use residential water and electricity bills but run a commercial enterprise can be identified according to the project and required to pay higher commercial rates. DDNS needs multiple stakeholders beyond the BMC—such as the water services company WATCO and the electricity company TPCOD —to use this system and has further plans to integrate other services like waste management. DDNS will also have monitoring dashboards and mobile applications for BMC officials to use.

22In the next section, I will show how the prevalence of consultants is expanding due to the neglect of existing state digital agencies, stunted municipal capacities with little training in the new digital infrastructure, and the promise of overcoming historical departmental silos and the lack of data sharing.

Implementing Digitalization: From State Capacities to Infrastructural Control

23All three platforms are presented as solutions, not by municipal agencies but by consultancy firms hired by the BDA and BSCL. The Bhubaneshwar Urban Knowledge Centre (BUKC) is based at the BDA, with a team intended to assist in town planning using specialized solutions for city planning, heritage management, and social media outreach, among others. The BUKC is led by the same consultancy firm that served as project manager for the Smart Cities program in the city. Despite being an offshoot of the BMC, the DDNS implementation team is entirely consult- and vendor-driven and does not have representatives from municipal agencies or even the municipal services companies. This is particularly visible in the case of Bhubaneshwar, where municipal elections were suspended for more than two terms, and the BDA and BMC were removed from public scrutiny. Even after elections were held in early 2022, the mayor and counselors had no fiscal or legislative powers to evaluate these “solutions.” Similarly, even though BLUIS is deployed through state agencies and has bureaucratic oversight, the platform and its algorithms are consultant-driven. In addition, although BPAS is touted as a state-level intervention by the state bureaucracy, it was developed by consultants for the smart city and implemented by a new vendor. In this case of weak municipal control, bureaucratic oversight offers some democratic supervision, but this is removed by the dependency on consultants and vendors. These are just three platforms among the myriad technologies implemented by the municipal agencies through these new institutions that are part of the Smart City and the Urban Knowledge Centre, setting the agenda for most spatial planning in the city without any visible citizen, political, or even bureaucratic engagement.

25In the next section, I will show how consultants continue to sideline existing state digital agencies, by introducing new platforms that offer “solutions,” such as “interoperability,” to address departmental silos and understaffing but burden untrained municipal capacities that cannot keep up with the new digital infrastructures.

Sidelining Existing Capacities and Agencies

Departmental Silos, Data Sharing, and Interoperability

28Some literature on the smart city has shown how public policy for digitalization strategies suggests that technological interventions and their datasets can be used to overcome the challenge of government departmental silos (Praharaj, Han, and Hawken 2018; Prasad, Alizadeh, and Dowling 2021). In the case of Bhubaneswar, it is clear that even this supposedly redeeming power of digitalization fails.

Untrained Capacities for Technical Infrastructures

31During my visit to the BDA offices over a period of two months in 2021–2022, I saw long lines of visitors seeking to understand why their BPAS applications were rejected. BDA officials had trouble retrieving applications on the portal or answering technical questions, rendering the claim of transparency moot. Recent news reports on BPAS mention user complaints ranging from lack of user-friendliness, a complicated file-entry system, OCAC server crashes, and multiple delays in approval (TNIA 2022). The BPAS statistics in Fig. 7 also show a significantly low rate of resolution despite the dashboards and escalation matrix. Twitter users and Google Play app reviews complain of a malfunctioning app, the inability to record pictures, and a lack of action despite complaints.

32Since platforms like DDNS and BPAS require working across municipal agencies, consultant teams like the ones at the BUKC or BSCL are given the power to implement these projects through “Request for Proposal” documents and contracts. The obligations laid out in the proposals and contracts are not public, and the viability of the solutions they offer falls outside the scope of any review. Similarly, once the contractual period for programming the platform ends, the actual implementation of these technologies is handed over to a vendor and later transferred under an Operations and Maintenance contract. Any problem with the implementation cannot be reviewed after the handover. The contractual nature of ideation-implementation-operation means that the “solution” offered by the firms is not under scrutiny. According to my multiple interviews in the BSCL and BDA, any failure of technology is often blamed on obligations between the vendor and the operator. The inability to find a solution is thus never the fault of the technology infrastructure itself, but is instead attributed to institutional failure, contractual mismanagement and infrastructure, or, ironically, on municipal capacities. Beyond the contractual justifications, none of the bureaucratic agencies have the technical capacity to evaluate how the platform and its algorithms function. The new technical platforms therefore add a layer of complexity with which neither institutions nor citizens are able to cope, leading to newer digital technologies that consultants offer as simplified solutions. An example of this is the replacement of the older BPAS developed by the BSCL with a new vendor that changed the code and the platform without questioning the platform’s ideation. As a result, this replacement did not change the recurring errors and delays.

34There are multiple signs that these municipal agencies cannot keep up with the new digital infrastructure, including understaffing, heavy workloads, and a lack of specialized capacity training for the new digital infrastructures.

Outcomes of Governance through Platforms

35While I have already identified some of the platform’s failings, I shall highlight the outcomes of these three ICT-GIS systems’ implementation. According to my findings, the shift to governance platforms results in reduced access to democratic institutions and processes, the transformative powers of algorithms in shaping laws and regulations, and an expanded scope for participatory surveillance.

Reduced Democratic Avenues

36The turn to digital technologies and platforms for municipal governance does not offer any avenues of conventional access to democratic institutions. A person whose potentially legal construction is voided by BLUIS or whose building plan is rejected by BPAS has no option of turning to conventional municipal bureaucracy or local counselors, since they do not have the capacity to scrutinize automated decisions. Visitors to the BDA seeking recourse must pass through another technological hoop in that a visitor’s pass has to be issued with a facial-recognition camera at the entrance. Even then, officials are unable to solve their inquiries. If the cross-linking of services under DDNS is fully implemented, it is unclear whom citizens can approach for redressal if their household is flagged for nonpayment of services due to an error in the database. In a context whereby the BMC does not have planning powers, the BDA outsources auditing to the BUKC, and projects under the BSCL lie outside of public accountability, since these governance platforms are devoid of any scrutiny.

Impact of Algorithms

38There are three takeaways concerning algorithms in governance. First, automation does not necessarily mean an algorithm is making decisions within a black box (Kitchin 2011), since the rules that drive it can be existing laws and codified rules. Algorithms cannot make decisions that deviate from the coded rules. The second takeaway is that specific algorithmic technology can lead to changes in municipal rules. The adoption of governance technology can thus be seen as having the power to reform legislation. Third, the impact of algorithms therefore extends far beyond its purported black box, burdening institutional capabilities even further—as seen in the case of BLUIS.

Hierarchical and Participatory Surveillance

39While BLUIS and BPAS are aimed at reducing the dependency on human interventions, they both require human action, either for scrutiny when the system flags it for official review or escalation, through ground visits for final building approval, or for on-site surveys for encroachment. Governance platforms are supposed to ensure that tasks are performed more quickly. While files with human personnel may be pending beyond the stipulated time periods, automated platforms like BLUIS have an inbuilt escalation matrix, which ensures that files get transferred to supervisors if they are not resolved within the encoded time period. If encroachment is not dealt with within a seven-day period, the platform escalates the matter to a zonal officer and so on until it reaches the Chief Minister’s Office (CMO). Escalation matrices and dashboards result in hierarchical surveillance that pressures lower-ranking officers to take action faster. In interviews, multiple officers within the BDA and the BMC often refer to phone calls from the “CMO” if files remain pending beyond stipulated time periods.

40Aside from internal institutional surveillance ensuring that tasks are completed and putting pressure on municipal officers, all three technologies create mechanisms for monitoring and surveillance in the city. While automating surveillance through satellite imagery, BLUIS also encourages citizens to act as agents in participatory surveillance to monitor public land. BPAS adds to the formalization of landholding records of the city while also monitoring the residential and commercial expansion of the city through its dashboards. DDNS promotes the municipal scrutinization of service consumption, economic activity, and the creation of a household-level database of residents that can be obtained by any of the partner agencies. “Before, the ward officer was like the neighborhood tax collector, but this is now difficult in a dense city … . [N]ow a dashboard can provide all household-level activities and the revenue that is obtainable,” said the DDNS consultant.

41In practice, platforms thus tend to reject to older democratic options and avenues, and algorithms have the potential to streamline legislation, even as capacities struggle to cope with its functionality. Meanwhile, the platforms are designed to encourage hierarchical and participatory surveillance.

Conclusion

42As the case of Bhubaneswar shows, the inevitable pathways for the digitalization of urban governance in India are not only set out through grand projects and policies like the Smart Cities Mission and the National Urban Digital Mission, but also through state- and city-level projects. While these pathways reflect the national tendency to rely on external consultants and vendors as sources of digitalization, urban administrative agencies in Bhubaneswar are also motivated when it comes to seeking platform solutions for urban-governance agendas.

43Contemporary platforms depend on scattered historical-digitalization strategies that have at times relied on either setting up state capacities, state institutions, and infrastructures or industry trends. While state capacities have declined and institutions have been sidelined, newer digital infrastructures have been designed to incorporate older databases or processes and follow current standards. Despite this, newer platforms show similar patterns of failure and dysfunction related to the difficulties of interoperability with legacy data, a lack of technical supervision or knowledge within municipal agencies, and a lack of data sharing, which are the hallmarks of municipal departments. However, the core issue is understaffing and a lack of capacity training within municipal agencies, which are unable to meet the expectations of the speedier and efficient resolutions that these platforms promise.

44In practice, the turn to platforms transforms the nature of urban governance. It does so by taking away redressal powers from democratic avenues and situating them further within technological solutionism. It suggests that algorithms are substantive enough to transfer most decision-making to platforms and modify existing governance rules to fit with these platforms. At the same time, the staff that are supposed to supervise or monitor these systems cannot understand the algorithmic decisions. The way these platforms are designed to incite participation, escalate matters, and send alerts and notifications increases surveillance within municipal hierarchies, allows municipalities access to household-level data, and promotes participatory surveillance by citizens.