VacSeen: A Suite of Linked Data-Based Applications for Improving Last Mile Vaccine Logistics in Developing Countries


Vaccine coverage in developing countries is subpar at about 80%. A key
reason for the subdued coverage rate is lack of complete, connected,
accurate, and timely data to inform logistical planning at the last
mile. On the other hand, the expansion of mHealth services in
resource-constrained settings presents an unprecedented opportunity to
electronically track vaccine administration in the field and detect
non-compliance. The availability of recipient-level data from the field
on one hand and semantically enriched data about vaccines on the other
creates an opportunity to relationally map and analyze them. By
exploiting an integrated set of proprietary and open data, governments,
health authorities, non-profits, and manufacturers can be empowered to
make informed decisions so that the right vaccine reaches the right
child at the right time.

Exemplars of such datasets include:

i. Data about socio-economic indicators and infrastructure in different vaccine administration provinces
ii. Linked Open Data about infrastructure set-up (such as airports, hospitals, and rail stations).
iii. Linked Open Data about known adverse events of vaccines
iv. CIA Factbook about data on vaccine administration sites
v. Data about road network and power infrastructure in the vaccine administration sites

Unfortunately, a lot of this data sits in isolated pockets and thus does not get utilized for decision-making. In this project, we are linking these datasets to help
the decision-makers in ensuring continued supply of vaccines to remote
regions of the world. VacSeen ( is a Linked Data-based, decision supporting platform for connecting myriad proprietary (logistical database from implemeting organizations) and open (ontologoies, weather, product information) data to enable more informed decision making with respect to immunization. The project currently features the following applications:

1. An ontology ( classifier to determine state of immunization coverage of Beninese vaccine recipients based on the country's national immunization schedule

2. Linkage of immunization information to product details such as brand names, adverse events, and biomedical identifiers

3. Multi-level authentication of scans of barcodes on packaages

In addition to these features, additional applications such as network-analysis based recipient mobility characterization have also been developed.

The project is being jointly implemented with VaxTrac, a U.S.-based non-profit with the aim to apply the VacSeen platform on immunization schedule datasets from field clinics in Benin (Africa).

Monika Solanki and Partha S Bhattacharjee

University of Oxford, UK and AutoID Labs, MIT
Wolfson Building, Parks Road
OX1 3QD Oxford
United Kingdom