Editor’s Note: This blog was authored by Susan Davis, executive director of Improve International, an organization focused on promoting and facilitating independent evaluations of WASH programs to help the sector improve. She has more than 13 years of experience in international development and has evaluated WASH and other programs in 16 developing countries. Her first career (8 years in environmental consulting) involved projects like combining databases across the 10 U.S. Environmental Protection Agency Regional offices, which is where her respect for unique identifiers was born. A version of this post originally appeared here.
What is a unique identifier?
You probably don’t think of it, but you use unique identifiers every day. In the U.S., your social security number is your unique identifier for the government (which is why if someone has it they can steal your identity). Your bank account number helps the bank track all information associated with you.
What is a physical unique identifier?
Well, your house has one — in the form of an address. Your car has one — the vehicle identification number. (The license plate might count but it is too easy to remove.) My dog has an identification chip embedded between her shoulder blades because her license tag could easily come off with her collar. A physical unique identifier needs to be permanent — long lasting in tough conditions, and not easily removable.
What does this have to do with water supply points?
The good news is that many more governments and NGOs are working to create inventories and to monitor services more rigorously. While there are occasional, limited efforts to include unique identifiers on water points, this is not a widely spread practice in the sector. Most plaques I’ve seen simply identify the donor and perhaps the date of construction. Currently most water points are named in reports and databases by the village or town in which they are located. This is not a reliable way of identifying water points uniquely. First, many villages have multiple water points, installed, rehabilitated, and/or replaced as they fail. Secondly, community names are often spelled differently in the indigenous language, and especially in English. For example, in Ethiopia English place names are often spelled phonetically (e.g., Gonder, Gondar). In Central America several villages have the same saint names. Thirdly, water systems vary from simple hand dug wells to complex spring fed gravity systems with several shared water points to pumped and piped systems with household taps. GPS capability on handheld devices is becoming more and more available, and several tools use it to help with water point mapping. However, it is not exact. According to GPS-basics.com, specifications for many GPS receivers indicate their accuracy will be within about 10 to 50 feet (3 to 15 meters), 95% of the time. This assumes the receiver has a clear view of the sky and has finished acquiring satellites. With consumer grade devices, we can usually expect to be within about 20 to 30 feet of the mark with most consumer grade receivers. The numbers can vary slightly and thus GPS-generated latitude & longitude can’t serve as unique identifiers in a database of water points. Others have suggested using photographs to uniquely identify water points. While a human might be able to match data that way, photos can’t be used by software programs to merge large amounts of data.
How will physical unique identifiers help improve sustainable services?
One of the most obvious ways is that a unique identifier on a water point would enable customers to report faults by calling or sending text messages to a mechanic or the responsible entity (M4water is trying this in Uganda; and Watertracker Ushahidi has a technical assistance system).
A rich database on water points would be a powerful and necessary tool to help governments, implementing organizations, and customers fully understand and address challenges to sustainable services. Currently, monitoring data on water points are collected by different groups, with different goals and indicators, and saved in different places. Data collected over time, even in the same area, only leads to unconnected snapshots and can’t be easily compiled into one database for analysis. Thus, water data are highly fragmented. Water quality, functionality, access, fee, and other data are collected from water points by different groups, including:
- National government inventories (e.g., Ethiopia, Liberia, Sierra Leone)
- NGO/UN organizations
- Local governments
A unique identifier physically applied to each water point would allow:
- Asset management at the national level
- More efficient monitoring
- Tracking of maintenance, repair, and replacements over time (along with associated costs)
- Community reporting (without needing GPS or even cell phones)
- Data layering for richer analysis — e.g., with health, population, income, water risk data
- Data comparison over time
How should these identifiers be generated and applied?
Some smart people have been thinking about this: see Akvo FLOW on ways to update data over time and mWater on Globally-Unique, Human-Readable Identification of Water Sources. Some data collection tools can generate a unique identifier, and I’ve heard suggestions for bar codes, RFIDs, or QR codes. But I keep thinking about the customer, and the local government. Will they have smartphones with barcode readers handy? What about the local NGOs who might be working with these communities? Below I suggest a few overarching guiding principles (I will leave the technical principles to the data experts):
- Keep customers in mind
- Pilot this effort in countries where national water point inventories are already established or underway
- In other countries, work with governments/national WASH networks to establish a scheme
- Keep it simple
I recommend that physical unique identifier systems be budgeted into all future water grants. To label all existing water points, budgets should be included in all national and organizational water point mapping efforts.
A community performs water quality tests monthly on a water point, a local government agency performs water quality tests on the water point annually, and an external organization verifies the results occasionally. All of that information can be easily compiled for the same water point. This would allow us to look at whether quality is improving or eroding over time, whether certain tests are more or less accurate, and so on.