Delhi’s mechanised road sweeping programme is falling short not simply because the city has too few machines, but because its road network, operating targets and street-level conditions do not yet align with the pollution-control task assigned to it. An assessment by the Commission for Air Quality Management, or CAQM, found that mechanised road sweeping machines clean only 35-40% of the capital’s large and medium-sized roads on alternate days, even though the prescribed norm is to cover the entire road network.
The gap is visible in the numbers. Delhi has nearly 100 mechanised road sweeping machines against a stated requirement of 172. The machines are expected to lift road dust before it is resuspended by traffic and wind, but their actual reach is constrained by road width, vehicle parking, encroachments, potholes and unpaved surfaces. The result is a system in which the formal cleaning target is much larger than the network that can be serviced consistently.
This matters because road dust is not a minor or seasonal issue in Delhi’s air-quality problem. A 2018 joint study by The Energy and Resources Institute and the Automotive Research Association of India, cited in the report, estimated that dust accounted for 42% of PM10 pollution and 38% of PM2.5 pollution in Delhi during summer. In winter, its contribution was lower but still substantial, at 25% of PM10 and 17% of PM2.5 pollution.
The figures do not mean that mechanised sweeping alone can resolve Delhi’s pollution crisis. They show instead why the city’s road-maintenance and street-management systems have become central to air-quality administration. A machine cannot deliver its intended pollution-control benefit if it cannot reach the kerb, if road surfaces are broken, or if parked vehicles and encroachments block the cleaning path.
Delhi’s network also creates a measurement problem. A senior CAQM official said the machines are used to clean roads with a combined length of about 15,000 kilometres. But a four-lane road has to be cleaned four times to cover all its lanes, increasing the total distance that must be swept each day. This means that a headline figure based on road length can conceal the operational scale of the task. The relevant question is not only how many kilometres of road exist, but how many running kilometres of lane must be covered, at what frequency and under what physical conditions.
CAQM’s own operating norms recognise this distinction. In February, the commission specified that large and medium-sized machines should cover about 40 running kilometres during an eight-hour shift, while small machines should cover 20 running kilometres in the same period. The reported operational performance is lower: a machine cleans up to 30 kilometres a day, compared with the 40-kilometre requirement cited by the official.
That shortfall suggests that procurement numbers alone cannot be used as a measure of implementation. Even if the city reaches the stated requirement of 172 machines, the programme will not automatically cover the entire road network unless the machines can complete their scheduled routes. Shift duration, route planning, road access, traffic management and the condition of the carriageway all affect the distance actually cleaned.
The distribution of Delhi’s roads makes the challenge sharper. Around 70% of the city’s total road network consists of roads less than 10 metres wide, according to the CAQM official. These roads are currently swept manually. The remaining 30% comprises large and medium-sized roads, of which only 35-40% are cleaned by machines on alternate days.
This is not merely a choice between two cleaning methods. It reflects the way the city has divided responsibility between road types and operating systems. Mechanised sweeping has been concentrated on roads with adequate right of way, while narrow streets remain dependent on manual labour. The MCD official quoted in the report said machines are deployed on roads with a right of way greater than 60 feet, while the civic body is procuring 70 machines for roads with a right of way greater than 30 feet. Narrow roads are swept manually by a workforce of about 60,000 sanitation workers.
The arrangement leaves a large part of the urban street system outside the mechanised programme. It also shows why machine procurement and sanitation staffing should not be treated as competing solutions. The city has different road geometries and different access constraints. Larger machines may be more productive on open corridors, while smaller electric machines are being considered for narrower roads. Manual sweeping remains the current system for much of the network, but the evidence cited by experts indicates that it can resuspend more dust.
Anumita Roychowdhury, executive director for research and advocacy at the Centre for Science and Environment, said mechanised sweepers have a moderate impact because they primarily lift coarse particles. She also said they are more effective than manual sweeping, which causes greater resuspension of dust. Her assessment places a limit on what the machines can achieve: they can improve one part of the dust-control chain, but they are not a substitute for repairing roads, managing roadside activity or addressing unpaved surfaces.
The physical condition of the road is therefore an institutional issue as much as an engineering issue. Roychowdhury said machines are ineffective on roads with potholes or unpaved surfaces and should be used alongside a proper road redevelopment programme. This links air-quality management to the responsibilities of road-owning agencies, including the maintenance and resurfacing of streets.
It also raises a coordination problem. CAQM has repeatedly asked road-owning agencies to use mechanised sweepers to lift road dust and reduce resuspension. The MCD has its own deployment criteria and procurement plans, while the Public Works Department had not immediately responded to the report. The city’s road network is therefore being managed through multiple agencies, each with different assets, road jurisdictions and operational constraints.
For residents and traders, the street itself is where these administrative divisions become visible. Experts have called for arrangements with residents and traders to ensure that parked vehicles and other obstructions are removed during scheduled cleaning. CAQM’s February norms also require operating schedules to be declared in advance for public convenience. These provisions indicate that road sweeping is not only a back-end sanitation activity; it depends on public communication and compliance at the time and place of cleaning.
The procurement of 30 smaller electric-based machines approved by the Delhi government in September is intended to address part of this problem by cleaning narrower roads. The MCD’s plan to procure 70 machines for roads with a right of way above 30 feet points in the same direction. But the supplied evidence does not establish how quickly these machines will be deployed, which roads will receive priority, or whether their routes will be published and monitored against completed running kilometres.
That missing operational information is important. CAQM’s norms specify a measurable daily distance, but the assessment also shows that the city’s real performance depends on whether that distance can be achieved in actual streets. A machine that is available but blocked by illegal parking, unable to operate on a damaged surface or assigned to a route with repeated obstructions does not provide the same pollution-control benefit as a machine completing its scheduled coverage.
The data also complicates the idea of complete coverage. If 70% of Delhi’s roads are narrow and manually swept, and only 35-40% of the larger and medium-sized roads receive mechanised cleaning on alternate days, the prescribed norm of covering the entire network remains a system-wide target rather than an achieved condition. The evidence does not show that every road must be cleaned by a machine. It does show that the current combination of machine capacity, route productivity and road access does not meet the stated coverage expectation.
Delhi’s dust-control challenge is consequently a test of municipal execution rather than machine ownership alone. The city needs road-owning agencies to coordinate route planning, sanitation operations, parking enforcement, encroachment management and road repairs. The public also needs advance information about cleaning schedules, especially where vehicles and commercial activity routinely occupy the road edge.
The CAQM assessment confirms a substantial implementation gap: nearly 100 machines are operating against a stated requirement of 172, mechanised sweeping reaches only 35-40% of large and medium roads on alternate days, and machine productivity is reported below the prescribed 40 running kilometres per shift. What remains to be established is how agencies will measure completed coverage, publish route-level performance and extend cleaning to streets where road design and obstruction currently prevent mechanised access.
