What is a digital twin, and why pair it with a shared budget?
2 November 2026 · The Cybernetic Society · Goa Business Commons
The front page says an AI digital twin sits behind the Goa Business Commons. This note says what that means in plain words, where governments already use digital twins and shared budgets, what the studies found, and what we have taken from each.
A digital twin, plainly
A digital twin is a working model of a real thing, kept in step with it, that you can ask questions of before you act on the thing itself. The idea came from factories and jet engines; cities adopted it once 3D maps, sensors and open data made a whole town modellable. A twin exists to be simulated: a government can try a decision on the model, see what it does, and change its mind before any money, land or trust is spent on the real thing.
Real ones, in government hands
Singapore.Virtual Singapore, built by the National Research Foundation with Dassault Systèmes, is a 3D model of the whole island that agencies query: how much energy a given roof would capture from solar panels, how shaded a walkway is at pedestrian level, how people would move across a proposed bridge in Yuhua. (GovInsider)
Helsinki. The city built two twins of its Kalasatama district in a 2018–19 pilot and published both as open data. On top of them sits the Energy and Climate Atlas, a free browser tool with each building's simulated heating demand and solar and geothermal potential. (City of Helsinki, final report)
Herrenberg, Germany. A town of about 30,000 people. The Stuttgart supercomputing centre folded wind flow, traffic simulation and residents' own ratings of places into one model, then set it up in virtual reality at events around town; about a thousand residents walked through it, and the council used it to consult on a shopping-centre project. (HLRS; Dembski et al., Sustainability 12(6), 2020)
Hamburg. In 2016 the city had to find room for roughly 20,000 refugees. Over seven weeks about four hundred residents sat at an interactive table model of their districts and placed housing themselves. Around 160 sites were accepted by residents and 44 passed the city's legal checks; some were built. (MIT Media Lab; URBACT)
Varanasi and Ahmedabad. Under the Smart Cities Mission, Varanasi Smart City Limited commissioned a 3D twin of 160 square kilometres of the city, operational by 2024 and a SKOCH gold award winner. Ahmedabad Municipal Corporation has since contracted a ₹283 crore city-scale platform that will give 25 lakh properties a digital identity, under the World Bank-backed Gujarat Resilient Cities Partnership. The early use in Indian cities has mostly been property-tax mapping; flood and traffic planning were the stated aim. (GIM International; ThePrint; The Ken)
Ours is a twin of something smaller
The Commons does not model buildings. It models businesses: each one that registers, where it is, what it said it needs, how it leans on five trade-offs, and which of 770 Goa government schemes could carry part of what it needs, read against every village and town in the state. That last part is the public half, and anyone can open it: the digital twin of Goa's panchayats, where each local body shows the schemes it is eligible for now and the ones it could plausibly reach.
The question our twin answers is the one a city twin cannot: which things could several businesses buy or run together, and what would each cost here? An AI reads the needs, writes a draft proposal when two or more needs rhyme, prices it for Goa, and notes the scheme that could co-fund it. It never decides. A draft reaches the ballot only when a business leads it and another backs it, and every proposal on the board says whether a person or the twin wrote it.
Participatory budgeting, plainly
Participatory budgeting is a rule, not a technology: a defined sum of public money, and the people it is for decide how it is spent, in the open, with the result binding. It is thirty-seven years old and has been run in more than eleven thousand places. (PB World Atlas, 2019)
Porto Alegre, Brazil, 1989. Where it began. Neighbourhood assemblies set the order of capital works. Between 1988 and 1997 the share of households with sewer and water connections rose from 75 per cent to 98 per cent. The city later wound the programme down, which is its own lesson about depending on one administration. (World Bank case study)
Kerala, 1996. The People's Plan Campaign earmarked 35 to 40 per cent of the Ninth Five Year Plan outlay for projects drawn up by the state's 1,214 local governments, with grama sabhas and task forces doing the planning. It is still the largest exercise of its kind in India. (NITI Aayog DMEO; Participedia)
Pune, 2006. The first Indian city to run it as a yearly routine: the corporation invites suggestions each monsoon, ward offices collect them, the prabhag samiti decides, and ₹50 lakh per prabhag is set aside in a separate budget annexure. In 2012–13 a city of 3.5 million sent in six hundred suggestions, which is the honest half of the story. (CEE review, 2013)
Paris, 2014–2020. Five per cent of the city's investment budget, about half a billion euros over the mandate, voted on by residents project by project. (Ville de Paris)
New York City. Council members put a slice of their capital money to a district vote. In the 2015–16 cycle 67,000 residents voted across 28 districts and funded 132 projects worth $38 million. (NYC Council)
Scotland. In 2017 the government and the councils' association agreed that at least one per cent of local budgets would go through participatory budgeting. By 2023 councils reported 1.4 per cent. (COSLA)
What the studies found
Three results stand up across different authors and methods. All are from observational data, so they show association and dose, not a controlled experiment.
Money moves toward basics. Sónia Gonçalves followed Brazilian municipalities from 1990 to 2004. The ones that adopted participatory budgeting put a larger share of the same budget into sanitation and health, and their infant mortality fell by one to three deaths per thousand births, roughly 10 to 20 per cent of the 2004 rate. (World Development 53, 2014, 94–110; summary)
The longer it runs, the more it does. Michael Touchton and Brian Wampler, with twenty years of data on Brazil's largest cities, found participatory budgeting associated with higher health spending, more civil-society organisations and lower infant mortality, and the link strengthened the longer a programme had been in place. (Comparative Political Studies 47(10), 2014, 1442–1469; Boise State)
People who take part keep taking part. Matching New York's participatory-budgeting voters to the state voter file, a 2023 study in Political Behavior found they were about 8 percentage points more likely to vote in ordinary elections afterwards, with the biggest gains among younger, lower-income and Black residents. (Johnson, Carlson and Reynolds)
The failures are as instructive. Porto Alegre's programme depended on one party's hold on city hall. Pune's survives but draws few people, because asking for suggestions is not the same as giving anyone a vote on a sum they can see. The pattern in the evidence is that participation has to be real (a known sum, a binding vote), broad (not only the organised), and repeated.
What the Commons takes from this
A known sum, in public. The fund is small and the position is on the front page: paid, approved, asked for, remaining. Payments go to vendors directly and every rupee is on the live board.
A binding vote by the people the money is for. Businesses, one vote each. Points across proposals, allocation rules chosen for fairness rather than a bare majority (the Method of Equal Shares among them), and a plan that stands only if most of those who vote approve it.
A twin that drafts, so participation costs an owner ten minutes, not a committee. This is the piece none of the examples above had. Herrenberg and Hamburg used their models to let residents see; ours uses the model to write the first draft, from what businesses said they need and what schemes already exist, and then gets out of the way.
Repetition on a stated timeline. A first round in thirty days, a second at sixty if money remains, a final outcomes report from the businesses at six months. Dates are on the front page.
A rehearsal before the real thing. Before the first business registered we ran a thousand simulated ones through the same software. What broke and what we changed is written up.
For an official reading this
Everything here is open. The scheme catalogue behind the twin (770 Goa schemes, with the eligibility rules we read from them), the per-panchayat twin at /twin, the round's full export, and the allocation code are available on request, and the board itself needs no login. If a department, a panchayat or a municipal council wants to run a round of its own, with its own sum and its own members, the software is built for exactly that; write to us at cyberneticsutra.com.
Helsinki, Kalasatama Digital Twins final report (2019): hel.fi; Energy and Climate Atlas via Helsinki 3D
Herrenberg: Dembski, Wössner, Letzgus, Ruddat, Yamu, “Urban Digital Twins for Smart Cities and Citizens: The Case Study of Herrenberg, Germany”, Sustainability 12(6):2307, 2020; hlrs.de
Hamburg FindingPlaces: Noyman et al., “Finding Places: HCI Platform for Public Participation in Refugees’ Accommodation Process”, 2018, arXiv:1811.10123; urbact.eu
New York: council.nyc.gov; Johnson, Carlson and Reynolds, “Testing the Participation Hypothesis: Evidence from Participatory Budgeting”, Political Behavior 45, 2023