Gujarat is experimenting with a new way to improve public systems: placing students, young researchers and startup developers alongside police and transport officials to work on problems that are usually handled inside government departments. The projects range from artificial-intelligence tools for criminal investigations and low-cost narcotics-testing kits to commuter reporting systems, crowd forecasting and research into tribal dispute-settlement practices. Together, they suggest that the state is treating young technical and social researchers not merely as trainees, but as an extension of its problem-solving capacity.
That shift is significant because the challenges being addressed are not abstract technology exercises. They are embedded in everyday urban administration: checking the backgrounds of migrant workers in industrial areas, managing police deployments during elections and VIP visits, responding to faulty escalators and ticket gates, integrating transport data and interpreting large volumes of legacy records. The collaborations also reach beyond cities, into tribal areas where understanding local customs may affect how newly posted officers work with communities.
The most structured example is Project Spandan, a Rs 12-crore initiative that brings university students into assignments associated with policing and public administration. The programme has four tracks: university-led research in areas such as artificial intelligence, machine learning, facial recognition, big-data analytics, geographic information systems, cybercrime and cryptocurrency forensics; internships for developing police applications and dashboards; training for police personnel; and advanced courses for selected officers at premier institutions.
Its design reveals an important institutional ambition. Instead of treating technology as a procurement item that can be bought from an outside vendor, Project Spandan connects research, training and operational use. Students and researchers work on specific departmental needs, while police personnel receive training in cybersecurity, law, behavioural analysis and digital evidence preservation. The model is intended to create a feedback loop between the people who use public systems and those who build them.
In Ahmedabad Rural, Superintendent of Police Om Prakash Jat has worked with young volunteers for two years. A 28-member Gen Z team has developed ADR-SHIELD, a collection of police portals and databases. One module records migrant workers arriving in industrial areas such as Sanand and Dholera and is intended to help police check criminal backgrounds. Other modules manage crime reports, personnel details, dossiers and election arrangements. During elections, the volunteers compile records on polling booths, officers on duty and presiding officers’ contact details. They also assist with route mapping and police deployment for VIP visits.
The significance of this work lies less in the novelty of a database than in the attempt to organise information that is spread across routine administrative processes. Police operations depend on timely access to records, but the report indicates that young researchers have also had to tackle basic data problems, including converting scanned legacy first information reports into structured information and translating Gujarati text into English. The difficulty is therefore not simply writing an artificial-intelligence model. It is making older, fragmented and locally produced records usable without losing their meaning.
A proposed AI-assisted investigation tool being developed with researchers from BITS Pilani illustrates the same point. The tool is being built using a decade of FIRs, chargesheets and Supreme Court judgements. It is intended to direct investigating officers towards relevant legal provisions and flag procedural errors before cases reach court. The dashboard is being designed to process records on local systems rather than send sensitive FIRs and chargesheets to public AI platforms, while making suggestions available in Gujarati.
Those design choices show how the public-sector use of AI is shaped by institutional constraints. Sensitive records cannot simply be moved to an open platform, and a tool designed for local officers must work across languages and document formats. The supplied report does not establish how accurate the proposed system is, whether it has been deployed operationally or how officers will review its suggestions. It does establish, however, that the project is trying to address the less visible work required before AI can be used responsibly: digitisation, language processing, local hosting and legal structuring.
Gujarat’s Metro system has provided a different setting for this experimentation. At Metro Hacks, organised by Google Developer Group Gandhinagar and GDG Cloud Gandhinagar, more than 200 developers travelled on a train between Mahatma Mandir and APMC while working on 13 operational problems identified by Gujarat Metro Rail Corporation officials. Participants had the duration of the two-hour journey to develop prototypes.
Several of the ideas focused on the relationship between commuters and control rooms. One proposed system would allow passengers to report faulty escalators and ticket gates by uploading photographs, with complaints feeding into the control room’s morning briefing. Other participants worked on an artificial-intelligence agent and a machine-learning system to anticipate stadium crowds and maintenance needs. Another project sought to combine Metro, BRTS and AMTS data for operational briefings.
This is a distinct urban problem from policing, but the underlying method is similar. The authority supplies operational problems; young developers build possible responses; and the exercise takes place close to the infrastructure being discussed. Holding the hackathon on a moving train also made the event an encounter with the transport system rather than a conventional coding competition. Participants were working around the realities of passenger movement, station operations and last-mile connectivity, even if the report does not establish that their prototypes have entered service.
The numbers indicate the scale of interest in this approach. Around 60% of Metro Hacks participants were students. At Sentinel, a CCTV integration hackathon held at Trinetra Central Command in Gandhinagar, more than 6,300 participants from 17 states took part, with a prize pool of Rs 51 lakh. More than 1,000 submissions addressed suspicious-activity detection, automatic number plate recognition, cross-camera vehicle tracking and watchlist matching. Participants also worked on mapping models, connecting departmental systems and calculating hardware and software costs.
These figures show that government-linked problem-solving events can attract a much larger technical community than a single department could employ directly. But participation is not the same as institutional capacity. A competition can generate many ideas quickly; a public system must be maintained, audited, secured and used consistently over time. The report says Sentinel was nearing its final phase when it was published, but it does not provide evidence on which submissions will be adopted, how they will be evaluated or what procurement and maintenance arrangements will follow.
The public-systems experiment also includes work that is not primarily technological. In Danta taluka of Banaskantha, three researchers from Tata Institute of Social Sciences spent six months documenting tribal customs and dispute-settlement practices. One practice, described as chadotaru, can delay the last rites of a deceased person until a dispute is resolved. The study was initiated by Suman Nala, superintendent of police for technical services, who had earlier helped bring back 29 families exiled under community social codes.
This part of Project Spandan widens the definition of innovation. For an officer newly posted to a region, understanding local institutions and social practices may be as important as having a new dashboard. The research is intended to help officers understand the area better, while giving young scholars a perspective on its social dynamics. It places local knowledge alongside artificial intelligence, GIS mapping and digital evidence as part of the information required to deliver public services.
There are also more immediate examples of youth participation in public safety. During Ahmedabad’s 149th Rath Yatra in July, Deputy Commissioner of Police Rupal Solanki brought in 149 young volunteers after an elephant had gone on a rampage during the 2025 procession. The volunteers monitored loudspeaker noise near elephants and shared live updates through a WhatsApp group. They also proposed a public discussion at Kavi Dalpatram Chowk. The exercise shows how young participants can be used for observation, communication and community engagement during large public events, rather than only for software development.
For urban governance, the institutional question is whether these collaborations can move beyond episodic events. A volunteer team, a hackathon or a research partnership can help an agency see a problem differently. Turning that insight into a dependable service requires decisions about data ownership, privacy, cybersecurity, language access, training, budgets and responsibility for failures. The report provides evidence of work in each of several problem areas, but it does not yet show a common framework for taking prototypes into routine operations.
That gap is particularly important for systems involving surveillance, criminal records and automated recommendations. CCTV integration, watchlist matching, cryptocurrency tracking and AI-assisted investigations can expand the reach of public agencies, but they also make the quality and governance of data more consequential. In the supplied material, officials describe the objectives and technical work, including local processing of sensitive police records. The report does not establish the safeguards, review procedures or performance results that will govern these systems once deployed.
The same implementation question applies to public transport. A commuter reporting application can shorten the distance between a fault and the control room, but its value depends on whether complaints are verified, prioritised and resolved. A system that combines Metro, BRTS and AMTS data could improve operational briefings, but the report does not say whether the different agencies currently share data in a common format or whether the prototype has been integrated into their workflows. The technical demonstration is therefore an opening step, not evidence of a completed service.
Gujarat’s approach is notable because it connects a large pool of young technical talent with the operational knowledge of public agencies. Ahmedabad Rural Police, Gujarat Police’s ATS, Gujarat Metro Rail Corporation, Trinetra Central Command, universities, IIT Gandhinagar, BITS Pilani, NFSU, TISS and startup-linked organisations are involved in different parts of the effort. The partnerships also allow students to work on subjects that formal courses may not cover fully, including hardware and software costing, departmental integration and the handling of legacy documents.
The model reflects a wider challenge facing cities and public institutions: demand for faster, more connected services is rising, while departmental systems often remain divided by function, format and institutional history. Young developers may bring new coding languages and technical methods, as officials in the report point out. But the government side contributes something equally important: access to real operational problems, domain expertise and the constraints under which public systems must function.
The evidence so far confirms an active pipeline of experiments rather than a completed transformation. Project Spandan has created a formal structure, police teams have developed portals and proposed investigation tools, Metro Hacks has produced prototypes, Sentinel has attracted thousands of participants, and researchers are documenting social practices relevant to policing. What remains uncertain is how many of these tools will become reliable, accountable and regularly used public systems. That outcome will determine whether Gujarat’s engagement with Gen Z becomes a durable administrative model or remains a collection of promising experiments.
