Our line of defense against Covid-19 are our hospitals which are critically exposed to infection since all patients converge towards entrances. Currently there are protocols in place to control entrance but they imply potential physical contact even using all protective measures. Howard steps in and allows gesture based triage on entrance and connectivity with medical staff. This approach also works for basically any other organization or establishment so it's a multi-purpose platform.


Team was creating a solution around touch-less technology and when Covid-19 hit EU we had long discussions what could Howard do to help, and then came #EUvsVirus where we understood that we can apply what we know and have as in-house-devised technology to address social distancing in general (or touch-less interaction) since that's the major infection vector. We realized that we all changed a lot as a community during this crisis and that there is no better time for global changes which can be facilitated with tech.

What it does

It prevents contamination in a contactless manner. Howard helps replace nurses and doctors actually doing the screening of patients on hospital entry for suspicion to covid-19 detection - thus removing a potential vector for compromising hospitals - our first line of defense against this threat. It can also screen anyone on any hospital entrance preventing unwanted contamination for an almost insignificant price per person (there are thousands of entries daily in EU's clinics/hospitals).

What is Howard

Howard is a modular server/client platform deployable in cloud, on premises or on single mobile devices, supporting currently a couple of edge modules:

  • interactive gesture-based touch-less triage
  • gesture-based mood barometer

Back-end supports centralized administration use-cases, configuration of "campaigns" and analytics. We focused in this hackaton on interactive touch-less triage as a first line of defense of medical offices in EU.

How we built it

Edge with gesture-based module running Covid-19 campaign

  • Edge front-end is running on node-js with on edge mongodb.
  • It's deployed on all-in-one machine (can be totally standalone and pre-installed without any internet/connectivity required, however that impairs capability to interaction with emergency/medical staff or any support channel)
  • It has a question-answer tree based on interactive bots which determine whether a patient has covid-19 symptoms (based on freely available medical symptom lists)
  • It was a notification of our gesture-based edge platform and time to develop it was minimal

Challenges we ran into

  • understanding all aspects of challenges when dealing with patients in triage
  • thinking we're not doing enough

Accomplishments that we're proud of

  • it works :D (as the videos show)
  • we demonstrated our Howard is really what we proclaim - a modular platform which can be easily modified to fit any case

What we learned

  • there's a large market potential for these kinds of solutions
  • gesture (video based) detection works and we should start using it
  • UX models for query-based types of solutions

What's next for Howard

  • light free technology demo version
  • approach mobile devices (android/iPhone/tablets of both sorts) and see how we can package
  • calculate scaling for diverse campaigns and multi-tenancy
  • product conception
  • first customers contracting
+ 2 more
Share this project:


posted an update

In a matter of 2 days our existing gesture recognition project got a new module for questionnaires as a PoC and it just works. It was a good opportunity to have exchange with experts from the field in Italy who challenged us to show what we aim and we got support and feedback that this can really help in the field.

Thus, looking back on the last evening of the Hackaton, congrats go to the team who made the new module, we proved what we preach :) kudos. Thanks also to our mentors who gave us the real-life experience from the field in most-hit areas of EU.

Log in or sign up for Devpost to join the conversation.