Dietary module: The dietary module estimates dietary exposure from pesticide residues in food or drinking water using algorithms consistent with EPA/OPP guidance. The dietary module looks just at the oral route.
Residential module: The CARES NG residential module estimates exposure to pesticides from products used in and around houses, such as insect repellents and indoor/outdoor foggers. The residential, module look at oral, inhalation and dermal routes.
The dietary and residential modules are based on the What we eat in America, N-HANES data. The residential module also includes CHAD data. The databases are publicly available. Our work was to analyse and harmonise this data before inclusion in the modules.
File management system: The file management system is similar to Windows Explorer or Google Drive. It allows the user manager their files. Overall the experience is elegant and imbued with emotional design principles. Our Users helped us in configuring the design to show what they needed in terms of meta data, filter and folder structures.
Data editor: Here the challenge was to build a mini Excel with features particular to data science. One of the unique features of the editor is a data validator for data cleaning and error management around data.
Wizard: The wizard is a stepwise interface that allows the User choose options and files that contribute to running a model. Once an assessment is run the job is passed over to a queueing system. This lead to some fun design around tracking the progress of a job and notifying the User on completion. The design required an understanding of the complicated workflow of the user. We ran numerous remote workshops building up to the design.
Collaboration: The cloud immediately makes collaboration an obvious step. CARES NG provides a shared folder where users can put assets they wish their team to have access to.
Testing and validation: We thoroughly tested both the dietary and the residential models by checking the calculations performed in multiple scenarios. The results obtained from assessments carried out with older software and CARES NG were compared to quantify the differences in the exposure level estimates due to the different models and underlying assumptions.
Documentation: We worked with our client on creating comprehensive online documentation for the CARES NG software and models. You can see this documentation here: support.caresng.org
Support: Design doesn’t stop at handing over the keys. We want to see that our product is used and want to know why if it isn’t. One step we took was to supply good sample data so that the user could run an assessment right away to learn the interface. We provided on-boarding that gives advice to the User when they arrive on certain pages of the application for the first time. Finally we use Intercom to capture instant feedback from the User on their concerns.
For more information visit CARESNG.org.