From small towns to smart cities.
A wish list for an open and affordable urban data platform for the developing world.
Wake up early in Bangalore or Hyderabad and you will invariably see the above scene. Large water tankers coming into posh neighborhoods, one after another, all belching diesel exhaust. They are most visible in summer when the city provides water for only two hours every other day. The water tanker suppliers suck the ground water table dry in surrounding farmland in order to bring water to the city. At Re. 1000 per 10,000 Lts, the water might be cheap for the above pictured apartments but for a significant chunk of these cities’ populace, it is a major expenditure.
This situation where municipalities fail to adequately provide essential services plays out in both big cities and small towns across a rapidly urbanizing India. Even with citizens paying for private services for garbage collection and sewage treatment, trash being burnt is a regular scene and lakes catching fire are nowadays too real to not be song lyrics anymore. These struggling municipalities are not only failing to provide equitable service to all their citizens but are also causing large scale environmental degradation.
Contrary to the popular assumption of wide spread ineptness and corruption, there are many municipal officials, engineers and workers with good intentions struggling with the outdated and under-designed infrastructure and ill-designed processes. Since they serve a vast populace (a typical city like Hyderabad or Bangalore will have about 7–8 million citizens with more than 8000KMs of water pipeline and sewage networks and about 5000 field workers for maintenance), they cannot interrupt service to make holistic changes and instead rely on ad-hoc and reactive solutions. This patchwork of fixes has accumulated over the decades and despite their best efforts, cities like Bangalore and Hyderabad now have Non Revenue Water (NRW) of about 30%–40% which means that 30%-40% of the water being supplied is either illegally tapped off, lost to leaks in the network or legally used but not paid for. Resources which are already in short supply are being wasted.
In order to tackle these urban challenges, there is a big push by the Indian central government to have a select few cities become Smart Cities and is helping the cities raise funds primarily for physical infrastructure (like water and sewage treatment plants) but also for information and communication technologies (ICT). The larger world-wide discussion on Smart Cities advocates layering technology onto city services for increased automation, remote monitoring, large scale data gathering and analytics and predictive modeling to improve public services with a heavy focus on data from sensors, IoT devices, cameras, etc.
However, quite a lot of the smaller cities and towns haven’t even transitioned from data gathering using pen and paper. The photos below are from the water department of a southern city of about 400,000 citizens and one which was selected to become a Smart City. A lot of data is meticulously collected but exists in a form which precludes data analysis and decision making. Needless to say, none of this data is available to citizens to ensure transparency and equitable service.
Even with outdated infrastructure and siloed primitive software systems, officials and workers in some large municipalities are imaginatively co-opting cheap smartphones for operational productivity. Some examples of such innovation are:
- GPS enabled smartphones and tablets for on-the-field updating of the GIS information for their waste collection, water and sewage networks, etc
- Cameras on smartphones used for ‘attendance selfies’, taking photos after clearing garbage bins, photographing water meter readings, etc
- Customers using smartphone apps to lodge complaints about potholes, water leaks, downed power lines, etc and also to send SMSes or call IVR phone lines to grievance redressal systems.
- Workers employing WhatsApp to communicate photos of maintenance work to keep their supervisors informed and for immediate spares/repair approvals for fixes
- GPS tracking of garbage trucks, water tankers and other municipal vehicles to curb diesel theft
- Apps which send SMSes to consumers to inform them of municipal services such as the schedules for availability of water, public restrooms, etc
Such innovation points to the adaptability of consumer technology and a desire to hack together productivity improvements. However, these apps and quick fixes are mostly developed by private contractors and in some cases by municipal staff in an ad-hoc fashion and are not fully integrated with the other ICT tools in the municipality. The following anecdote illustrates the ad-hoc nature of the mobile innovations.
In this photo, a municipal worker is unclogging a domestic sewage connection. (The clogs happen frequently because the rain runoff drains and sewage drains are not kept strictly separated.) The unclogging happens as part of the following sequence of events:
- The house owner called an IVR number to register his complaint and provided his/her water utility/connection number, phone number, location and a description of the problem.
- Most municipalities have a customer grievances or complaints system (of at least 10–15 years vintage) usually built with Microsoft/Oracle/SAP software and databases. The city’s geographical spread is broken down into divisions (or wards). Each morning, the manager in charge of the customer’s division downloads the list of complaints in his division.
- The maintenance workers make their rounds either on foot or bicycle so the manager and the team prioritize complaints in geographical order. They do it with their knowledge of the neighborhood and do not use any mapping tools. They rarely take along a print out of the GIS based maps of the pipeline or sewage network.
- The field worker goes to the neighborhood, calls the phone number listed on the complaint, figures out the correct address and goes to the location.
- The GIS maps can be unreliable if the network is old or if it was laid without proper GIS documentation. There might be some exploratory digging before the network is fully scoped out.
- If the problem needs specialized equipment (or is a leak requiring repairs to broken pipes, ferrules, etc) the field worker takes a picture with his phone (many own cheap Android smartphones) and sends it to his manager on WhatsApp for approval.
- After the repairs are approved, the worker collects the equipment/replacements from the stores and fixes the issue. They might take another ‘after’ photo and the manager eventually marks the complaint fixed.
- The citizen is sent an SMS or email informing him of his complaint getting closed and the online status for the complaint moves from ‘Pending’ to ‘Closed’.
This anecdote is from a large city with a budget for the automation of some parts of its network (using SCADA and flow/pressure/level meters/sensors), ERP systems, IVR systems, a GIS data department, mobile apps, etc. Despite all the ICT tools, the operations and maintenance folks cannot access and utilize the data in the field. For smaller cities and towns, most of the sequence of events would just be co-ordinated over phone with locations marked on print outs of the GIS maps. For network issues without a customer complaint attached, such as leaks in a bulk pipeline, unless a concerted effort is made, it is a long time before the issues are discovered and acted upon. In simpler words, city services data is not collected properly because there is no destination for the data and whatever data is collected, is not easily accessible because it is disorganized.
The anecdote perfectly illustrates the wish list for a smart urban data platform.
- A municipality provides its services over a large geographical area and the GIS information underlying its services should be the foundation for any user interface. The smart urban platform should be able to support easy imports from tools such as AutoCAD used extensively by municipalities.
- The processes, touch points and data models of City services such as water supply, solid waste and waste water management, etc should be standardized, open and documented. This would permit easy on-boarding of existing systems and allow for customization.
- The urban data platform should have distinct UIs serving 4 personas: decision maker (such as heatmaps across wards/divisions for the municipal commissioner), operations manager (such as network alerts for the water network engineer), field worker (such as location aware network maps for linemen or valvemen) and citizen (such as water supply schedules or public toilet locations for their neighborhood).
- The urban data platform should accomodate data streams from citizen feedback/complaints systems and location aware mobile field apps (can be simple survey/forms for data collection). These data streams are as crucial as streams from SCADA systems, sensors/IoT devices, etc.
- The platform should have clearly defined interfaces/APIs to allow integration of existing databases and ERP systems. Even a basic Excel upload helps onboard legacy data.
- The smart urban data platform should be built for secure cloud hosting to simplify scalability and administration.
- The platform should be open-source and well documented to allow for the third party (and preferably local) development of analytics, easy integration of third-party sensors/devices, customization and deployment.
- And finally, it should be intuitive and affordable for small towns to readily adopt it.
At SmartTerra, we are a team of technologists working closely with municipalities to build such a platform. Intended to be open-source and available for all municipalities. More posts to follow with details of the platform. Please stay tuned!