# Robotic EV Charging in India: Pantograph & Depot Automation

> Robotic arms, roof pantographs and floor contacts compared for Indian depots, with the standards, power limits and real costs that decide the case.

Canonical: https://roboticsengineering.in/robotic-ev-charging-automation-india/ | Published: 2026-10-31 | Updated: 2026-10-31 | Author: Jignesh Panchal

**Robotic EV charging only pays in a depot where the same vehicle returns to the same bay every night and a person has to walk a cable to it. In India that means bus depots first, then truck yards, then nothing else for a while. The mechanism you choose — an arm with a plug, a roof pantograph, or a floor contact — decides the civil drawing, the power connection and the safety case before a single charger is ordered. This guide separates the three, sets out what the standards now require, and gives you a brief you can send to suppliers this week.**

## Why depots come first

A public charger serves a vehicle whose arrival time you cannot predict. A depot charger serves a vehicle whose arrival time you know to the half hour: the bus is back at 22:00 and leaves at 05:30. Predictable dwell is the only condition under which automation of the connection pays, because the saving is not the thirty seconds a driver spends plugging in. The saving is the queue that forms when twenty buses return together and three chargers are free.

The money is pointed the same way. PM E-DRIVE carries a ₹10,900 crore outlay, with ₹2,000 crore earmarked for public charging infrastructure and ₹4,391 crore for 14,028 e-buses, of which 13,800 were allocated as of August 2026. Buses that size cannot be charged at a home socket, so depot charging is where the e-bus allocation actually lands. That is the demand signal a charger vendor or a systems integrator is responding to.

Everything else in this article assumes that precondition. If a site has two vehicles, four hours of dwell and a driver standing next to the vehicle anyway, automation adds cost and failure modes with no queue to remove.

## Three ways a machine can connect a vehicle

The three mechanisms are not variations on a theme. Each one changes the building, the vehicle specification and the maintenance contract.

| Mechanism | Typical power | Dwell it suits | Headline cost | Where it fails |
| --- | --- | --- | --- | --- |
| Robot arm with standard connector | 22 kW AC to 150 kW DC | 4 to 10 hours overnight | Arm, gripper and integration on top of the charger | Seat force on the connector, cable management, dust in the plug |
| Roof pantograph, contact from above | 150 to 600 kW | 5 to 20 minutes | ₹25 to 50 lakh per charger for 18 m e-buses | Roof space, height tolerance across models, fleet standardisation |
| Floor contact or underbody plate | 50 to 200 kW | Short opportunity stops | Civil works dominate | Water ingress, debris, maintenance access under the vehicle |

*Power and cost bands are indicative; the pantograph band is the figure used in 2026 government evaluation of 18 m e-buses.*

The arm is the only option that works with a vehicle already on the road today, because it uses the connector the vehicle already has. It is also the slowest to connect and the most sensitive to where the driver parked. The pantograph is the fastest and the most expensive per bay, and it only works if the whole fleet shares a roof interface. The floor contact is cheap per vehicle and expensive per building.

A fourth option exists and is often the right first step: a [mobile robot that carries the charger to the parked vehicle](https://roboticsengineering.in/mobile-robot-charging-docking/) instead of bringing the vehicle to a fixed charger. It suits yards where bays cannot be rewired and power is only available at the perimeter. The trade is a battery on the robot and a docking routine that has to be reliable in the rain.

## What the standards now require

Until this year, an automated connection device was an improvisation that a safety assessor had to reason about from first principles. That changed in June 2026 when IEC published 61851-23-1, which defines automated connection devices for DC charging in three system classes, A, B and C, and sets the requirements for each. The practical effect is that a supplier can now be asked which class it claims and which test house verified it, instead of describing the machine in prose.

The robot itself is governed separately. [ISO 10218](https://roboticsengineering.in/robot-safety-standards-india/)-1 and 10218-2 were revised and published in February 2025, the first major revision since 2011. The 2025 edition folds collaborative operation into the main standard, adds cybersecurity requirements where a network fault can create a safety fault, and reworks the functional safety clauses. If the plugging arm is an industrial robot, this is the document your integrator should be working to, not the 2011 text that still appears in a lot of tender documents.

India’s low-power standards remain relevant in mixed depots. Bharat AC-001 specifies three outputs of 3.3 kW each at 230 V single phase, and Bharat DC-001 specifies up to 15 kW at 200 A and 40 to 100 VDC, which covers two- and three-wheelers and light commercial vehicles. Neither is an automated connection standard, and a depot that mixes them with a high-power pantograph needs separate load accounting for each.

One caution that belongs in the brief rather than in a footnote: India adopts IEC charging standards through the IS 17017 series, but the precise part-by-part mapping and the publication dates of each part are not something a buyer should assume. Require the supplier to name the exact edition it complies with and the test report number. A supplier that cannot name the edition is describing a prototype.

## The power constraint nobody budgets for

Chargers are the visible cost. The connection is the one that stops the project. Karnataka’s tariff order sets a demand charge of ₹350 per kVA per month for EV charging in BMTC and KSRTC depots under LT-6(c). A depot with 1,000 kVA of sanctioned load pays ₹3.5 lakh every month before a single unit of energy is consumed, and a depot sized for twenty 150 kW bays at simultaneous peak needs 3 MW, which is ₹10.5 lakh a month in demand charges alone.

That arithmetic drives two design decisions. The first is load management: stagger the charging windows so the depot never draws its theoretical peak, and the sanctioned load can be smaller than the sum of the chargers. The second is the application timeline: a new 3 MW connection involves the distribution company, a transformer, and in many industrial areas a feeder upgrade, and it takes longer than the chargers take to arrive. Start the connection application before the charger purchase order, not after.

Energy charges sit on top of the demand charge and vary with the time-of-day slab. Overnight charging in a depot usually lands in the cheapest slab, which is another reason the depot model works and highway fast charging struggles: the vehicle is asleep exactly when the tariff is lowest.

## How the depot layout changes

A manual depot is drawn around the vehicle: bay width, turning circle, a driver’s door. An automated depot is drawn around the robot’s reach and the vehicle’s stopping tolerance. A driver will park within roughly a third of a metre of where you asked. An arm that has to insert a connector needs the vehicle within a few centimetres, which means either a mechanical guide, a floor marking plus a sensor, or a robot with enough compliance and vision to find the port. Decide which of those three you are buying, because it is the difference between a ₹20 lakh arm and a ₹20 lakh arm plus a positioning system.

The second change is where the machine lives. A gantry above the bay keeps the floor clear and puts the arm out of the path of a moving bus, at the cost of structural steel. A floor-mounted arm is cheaper and has to be protected by a bollard that a reversing bus will eventually hit. A column mount splits the difference. The layout rules that apply to any robot cell — reach envelope, exclusion zone, maintenance access — apply here with one addition: the vehicle moves, so the exclusion zone has to be enforced against a driver, not only against an operator. The method is set out in our [guide to robot cell layout design](https://roboticsengineering.in/robot-cell-layout-design/).

The third change is cable and connector management. A cable left on the floor is a trip hazard and a wear item; a cable on a retractor is a mechanism that fails in the rain. Pantographs avoid the cable entirely and pay for it in roof real estate and in the requirement that every bus in the fleet present the same contact geometry at the same height.

## What it costs in India

Pantograph chargers for 18 m e-buses are being evaluated by the government at ₹25 to 50 lakh per unit as of August 2026. That figure is for the charger, not for the bay: add roof structure or a gantry where the bus cannot reach a wall-mounted unit, add the transformer and the connection, and add the civil work. A twenty-bay depot with pantographs is therefore a ₹5 to 10 crore charger line item before the building changes.

An arm-based bay is cheaper on capital and more expensive on engineering. A six-axis industrial arm in the payload class needed to hold a DC connector and its cable is in the same range as the arms used for machine tending, which our [industrial robot cost guide](https://roboticsengineering.in/industrial-robot-cost-india/) breaks down by payload and brand. On top of the arm sit the gripper, the vision or guidance system, the safety devices, the integration and the acceptance testing. Integration is usually the largest single line and the one most often underestimated. Lead times for the arm itself are covered in our [lead-time guide](https://roboticsengineering.in/industrial-robot-lead-time-india/); the charger electronics often arrive sooner than the connection approval.

The comparison that decides the business case is not robot versus human. It is automated bay versus an extra shift of drivers and an extra charger. If a depot needs four people on a night shift to plug and unplug 60 buses, the wage bill, the attrition and the queue are the baseline. Automation has to beat that baseline, not the cost of one person with one cable.

## Safety: two systems that can hurt people

A 600 V DC system and a moving arm share a bay. The safety case has three parts. The first is the charger: isolation, insulation monitoring, arc fault detection and a connector that cannot be energised before it is fully seated. The second is the robot: ISO 10218-1:2025 compliance, a validated safety function that stops the arm when a person enters the zone, and an emergency stop reachable from the driver’s door. The third is the interface between them: an interlock that prevents the arm from moving while the connector is live, and prevents the charger from energising while the arm is in the bay.

Roof pantographs add a wet-weather problem. Water on a bus roof and a high-voltage contact in the same place means the insulation monitoring has to be trustworthy, the earth path has to be designed rather than assumed, and the contact has to be mechanically guarded against a person reaching it. Floor contacts add a different one: debris and standing water under a vehicle that nobody crawls under to inspect. Both are solvable, and both are solved on the drawing board, not in the operating procedure.

The fault that catches Indian depots most often is not electrical. It is a bus parked half a metre out of position, a driver who gets out to check, and an arm that was told the bay was clear. Vehicle-presence detection has to be independent of the parking guidance system, and the arm has to treat an unexpected presence as a stop, not a warning.

## How to brief a supplier

A brief that asks for “robotic charging” gets three incompatible proposals. A brief that asks the following ten questions gets proposals you can compare:

- Which automated connection device class under IEC 61851-23-1:2026 does the system claim, and which test house verified it?
- Which edition of ISO 10218-1 does the robot comply with, and what is the validated safety function?
- What vehicle-positioning tolerance does the system need, and what provides it?
- What is the tested connector insertion and withdrawal cycle count, with the test report?
- What happens when the sequence fails halfway: vehicle present, connector not seated, arm mid-travel?
- What is the charger topology, and how does it handle two vehicles on one power stack?
- Does it speak OCPP 2.0.1 to a depot management system, and can that system see the robot state?
- What is the service response time, and where is the spare connector held?
- What does the site survey cover: power, civil, vehicle sweep, roof height across the fleet?
- Which depots in India run this system today, and can we speak to the operator?

The last question is the one that separates a supplier from a demonstrator. A pilot in one depot with one vehicle type is not evidence that the system works across a fleet with three OEMs and three roof heights.

## Four mistakes that sink depot charging projects

First, choosing the mechanism after the civil drawing. The mechanism determines the bay, the structure and the power; a depot designed for cables and then fitted with pantographs pays twice.

Second, sizing the connection for simultaneous peak. Almost no depot needs every bay at full power at the same moment, and the demand charge is charged whether or not you use it.

Third, writing the brief around charger kilowatts instead of vehicle dwell time. The useful number is how many kWh a bus needs in the window it is parked, not the nameplate of the charger.

Fourth, ignoring fleet standardisation. A pantograph system is only as good as the least standardised bus in the fleet, and a mixed fleet bought over five years will have at least three roof interfaces.

## Where to go next

If the depot also assembles or services battery packs, the charging bay and the pack line share a power design and a safety culture; our guide to [EV battery pack assembly automation](https://roboticsengineering.in/ev-battery-pack-assembly-automation/) covers the pack side, and [battery management systems in robotics](https://roboticsengineering.in/battery-management-system-robotics/) explains what the vehicle will tell the charger about its state.

Once more than a handful of automated bays exist, the operating problem becomes scheduling and fault response across the site. A [robot fleet management system](https://roboticsengineering.in/robot-fleet-management-software/) is the layer that tells the arm which bay is free and the charger when to expect a vehicle; without it, each bay is an island.

To compare the automation budget against the rest of the plant, see our [warehouse automation cost guide](https://roboticsengineering.in/warehouse-automation-cost-india/), and before you write the specification, read [how to choose an industrial robot in India](https://roboticsengineering.in/how-to-choose-industrial-robot-india/) — the same selection logic applies to the arm that will hold the plug.

## Frequently asked questions

**Is robotic EV charging already in use in India?**

Partly. Pantograph charging, where a charger contacts the roof of the bus, is in depot use and is being evaluated by the government for 18 m e-buses at ₹25 to 50 lakh per charger as of August 2026. Robotic arms that insert a standard connector are at pilot and vendor-demonstration stage rather than routine depot operation, and the connector-insertion reliability data that would justify a fleet purchase is still thin.

**What does a pantograph charger cost in India?**

The figure used in 2026 government evaluation is ₹25 to 50 lakh per charger for 18 m e-buses. That excludes the transformer, the connection, the roof structure or gantry, and the civil work around the bay, which together can approach the charger cost at a site that needs a new connection.

**Which standard covers an automated charging connection?**

IEC 61851-23-1, published in June 2026, defines automated connection devices for DC charging in three system classes A, B and C. The robot arm itself is covered by ISO 10218-1:2025 and the cell by ISO 10218-2:2025, both published in February 2025. Ask a supplier for the edition and the test report, not a claim of compliance.

**How much power does an e-bus depot need?**

Multiply the number of bays by the charger power only if every bay charges at once. A depot with twenty 150 kW bays has a theoretical peak of 3 MW, which at the Karnataka LT-6(c) demand charge of ₹350 per kVA per month is ₹10.5 lakh a month before energy. Staggered charging usually cuts the connection substantially, and the connection application is the longest item on the schedule.

**Can a robot plug a CCS2 or Bharat DC-001 connector into a vehicle?**

Technically yes, and it is the only automated option that works with vehicles already on the road, because it uses the connector the vehicle already has. The engineering difficulty is insertion force and alignment: the arm needs the vehicle within a few centimetres, and the connector has a finite cycle life that a depot will consume quickly. Ask for a tested cycle count with a report before accepting a reliability claim.

**Is automated charging cheaper than hiring more drivers?**

Not on labour alone. The case rests on the queue: if a depot cannot plug sixty buses fast enough within the overnight window, automation buys throughput that an extra driver cannot. If the queue does not exist, the capital cost of an automated bay will not be recovered from the wages of one person with one cable.

## Sources and further reading

- [IEC 61851-23-1:2026, Electric vehicle conductive charging system, Part 23-1: DC charging with automated connection device](https://webstore.iec.ch/en/publication/92838), IEC Webstore, published June 2026.
- [PM E-DRIVE allocation for e-buses and public charging](https://www.pib.gov.in/PressReleasePage.aspx?PRID=2294482), Press Information Bureau, August 2026.
- [PM E-DRIVE scheme portal](https://pmedrive.heavyindustries.gov.in/), Ministry of Heavy Industries.
- [Standardization of protocol for EV chargers, Bharat AC-001 and DC-001](https://heavyindustries.gov.in/sites/default/files/2023-09/Standardization%20of%20protocol.pdf), Ministry of Heavy Industries.
- [Bharat AC-001 charging station specification](https://bel-india.in/product/ev-charging-station-model-no-bharat-ac001/), Bharat Electronics.
- [Karnataka Electricity Regulatory Commission retail supply tariff](https://kerc.karnataka.gov.in/), LT-6(c) tariff for EV charging in BMTC and KSRTC depots.
- [Updated ISO 10218: answers to frequently asked questions](https://www.automate.org/robotics/blogs/updated-iso-10218-faq), Association for Advancing Automation.