New updates and improvements to Ouva.
Our new architecture uses a variety of models and methods working in parallel to get better at identifying patient activities. We are constantly making improvements, so keep an eye out for further updates as we build on this new platform.
We've implemented a major update to staff visit recording. New system ensures staff visits are recorded accurately as one activity when there are temporary interruptions in detection.
We updated text and language around the application, particularly in the notifications and patient activity items to be more descriptive.
This update brings an update to patient overview, system fixes and API changes.
We updated the patient overview page layout to emphasize the room status and the current alerts. In subsequent updates, live view will be only available to system administrators.
We have reduced bandwidth and processing power requirements, introduced several new user features and continue to make Ouva the best healthcare enterprise AI solution for virtual care.
We decreased the minimum recommended resolution across all sensors by half. This reduces network bandwidth and processing power loads needed.
You can filter and sort the dashboard dynamically. For example, you can sort rooms by priority, first showing the rooms with the most critical alerts.
You can filter patient and configuration tables. Click on the 🔍 icon found next to column titles to search.
The inbox indicator dot now shows the most critical alert color. You can find it on the side menu when the inbox is closed.
We have highly optimized our AI models for NVIDIA processors. This new upgrade now allows for lower latency and higher throughput on the Ouva cloud and your data centers that NVIDIA powers.
We replaced the previous pose-based people detection model with a new modern one that is faster and performs better in cases where the system previously failed (e.g., sitting in bed with the back towards the camera).
As part of this new change, the live view no longer displays poses and shouldn't be relied on to understand patient status. With this update, the Ouva bed and patient status are more accurate, and users should use it instead of live visualization. We are currently soliciting feedback from customers and partners to deliver an improved live view to solve specific telehealth challenges.
We implemented a new 'Beds' dialog, where sensors and beds can be created together for fixed camera deployments. With the 'Create more' button, you can quickly set up as many devices as you need without reopening the dialog.
These technical changes are related to security and deployment procedures.
In this update, we have made several quality-of-use improvements, increased Ouva performance and released some changes to the API for integration partners.
We are beyond thrilled to announce the launch of our version 1.0 after four years of research and development followed by beta launch in 2022. This update also features a brand-new dashboard view, designed meticulously based on the feedback from nurses, to provide a clear visualization of data from the bedside.
This update brings new changes enhancing data interperatability, user experience, improving system stability, and bolstering backend architecture for better scalability and redundancy.
In this update, we have brought quality of use improvements to daily use of the Ouva platform.
We have taken the feedback and made the charts more readable, ensuring clear visual distinction between different activity types and easier data interpretation.
All ongoing events (e.g., bed leave alerts, bed turnover delays) are now pinned to the top of the activity log. This way, even if a warning has been active for a long time, you will see it when you visit the patient.
.jpeg)
In this update, we have made several under-the-hood changes to the Ouva platform to reduce issues and improve usability.
We are excited to announce the release of Ouva version 0.13, which includes several new features and improvements to enhance patient observation capabilities.
We have completely revamped our AI for capturing patient activity, built on years of research. We trained a completely new model, which now uses thousands of images collected over several years. We made it more robust and accurate, going beyond pose capture and allowing for the addition of new capabilities quickly. Furthermore, the updated model requires fewer hardware resources, making it even easier to scale AI across hospitals.
Our upgraded room capture AI has increased performance and significantly higher accuracy. It automatically detects equipment such as beds and chairs, eliminating manual calibration and making it easier for your nurses to focus on patient care.
We have implemented optimistic updates in the web app for the department, notification, and patient records. This new method applies your changes immediately, improving the speed and efficiency of your interaction with the Ouva dashboard.
We have also fixed a bug that caused disconnection of the WebSocket on room transfers, ensuring a smoother experience for your caregivers.
In this minor update, we have improved the following:
In this update, we've added fall detection support, improved mobility widgets and API, enhanced app responsiveness and tracking performance, updated inbox UI, and fixed numerous bugs related to notifications, request handling, and UI display.
In this update, we've focused on enhancing the overall user experience by fixing various frontend and backend bugs, improving security measures, and optimizing performance. Key updates include:
.png)
To display data from the bedside within milliseconds on every device, we now load only recent patient activity when viewing a patient. We moved all data over 12 hours old to a separate view, where you can explore the entire patient log.
In this minor update, we have resolved several minor but persistent issues.
.png)
As we move towards the official 1.0 launch, this update brings inbox, activity log, updated design, new events and more to nurses and sitters everywhere.
We have launched a new way for nurses and sitters to access their departments' most important alerts and warnings. Through Inbox, you can get fall or wander alerts, round and patient flow delays detected via AI and delivered to you no matter where you are within the app.
If you are in dashboard mode (displaying all patients on a screen in a hallway, for example), you can hide the Inbox. Clicking on an alert will take you directly to the patient, where you can see their status and whether they are alone.
Suppose you think the alert was thrown in error. In that case, you can manually resolve it, informing our team to improve the detection models.
Suppose a critical alert disappears before you see it (for example, a fall-risk patient leaves the bed and gets back). In that case, you can find it under Recently Resolved. And if you would like to continue to alert the team, you can Unresolve it.
You can now find all previous alerts and warnings, times when people and staff arrived and left the bedside, patient's mobility events, and more under the Activity log. The ongoing alerts (such as an active fall risk) and events (discharged patients waiting to be taken out of the room) are shown on top until they are resolved.
When you view the log, a line separates all previous events from the ones you have not seen yet, allowing you to quickly catch up with the bedside activity.
We have updated everything about how the Ouva application looks and feels with the primary goal of easier access and better visibility of immediate actions available.
To show only the relevant content and not overload the views with data, we have separated patient charts into their own tabs. We created a section on the Overview page where you can see the most essential patient chart relevant to your department.
We now display an ongoing event from when a patient order (admit, transfer, discharge) is received until its completed. For example, when a bed is assigned to a patient, an ongoing event shows that Ouva is awaiting the patient's arrival. If there is a delay based on a department-level threshold, it displays a warning accessible across all devices. When the patient arrives at the room, the event is marked complete automatically.
We now support multi-level alerts. For example, if a fall-risk patient sits on the bed, you will receive a warning. The notification will turn into an alert when the patient attempts to leave the bed. When the patient sits back, or a nurse or visitor attends to the patient, the notification automatically resolves. All the settings are configurable based on the department's needs.
If your telehealth cameras run on WebRTC, a common video communication protocol, you can enter its link into the device settings and have its stream processed by Ouva just like any other camera.
This update includes changes behind the scenes that improves overall reliability, accuracy and safety of the Ouva platform.
.png)
Ouva offers time-critical applications, and for some, every second counts. After listening to customer feedback, we drastically improved the patient page's loading time by adopting a layered approach to how we load data so that you can get the information as it appears instead of waiting for the entire page to load.
You can now see when a nurse or doctor last visited the patient. When a visit is in progress, you can see how long it has been.
We made several performance improvements to the entire platform.
We have made several improvements to secure deployments further.
We have added new alerts from the services running on the Ouva platform. These notifications will allow our engineering team to respond to unforeseen issues rapidly.
Assistant displays can now show floor directory alongside the map. Locations are automatically grouped by the direction from where the patient is currently looking at the screen and commonly recognized destinations will show their respective icons.
.png)
You can now access our support team via Intercom integration live by clicking on the chat button on the bottom right.
Improving patient mobility reduces the time to recovery. Ouva non-invasively tracks patient ambulation and reports to care providers on their progress. This update has brought new data analysis capabilities to understand and act on this data.
Mobility programs in healthcare change from department to department and across patient conditions. You can now set department-wide or per patient mobility targets. Ouva tracks patient progress towards this target daily.
Understanding patients' mobility trends is an integral part of a recovery program. Ouva now displays daily and weekly charts to show how patients progress towards their discharge targets. When they take over a shift, nurses can glance at all patients to see who needs the most attention.
Ouva now integrates with the hospital ADT feeds to automatically receive the latest patient and bed requests. Ouva compares the requests to the actual live data it gathers from the bedside to predict and prevent patient flow delays.
We have revamped our camera integration to support cloud telemonitoring applications. Now, running Ouva is as simple as entering a link to a camera web stream.
We have moved our entire API to the more efficient and flexible GraphQL. As one of the industry-standards, GraphQL is a modern framework that allows building complex and quick interactions.
We released a brand new patient history page that displays all data from the previous patient stays. Administrators can identify past issues (e.g., falls, discharge delays) by reviewing the timeline of all events from all the beds patient has been admitted and transferred.
As we continue to add more insights by working together with nurses and physicians, here are a few things you can do today:
Ouva now detects chair sitting as a separate activity. You can see the new activity in the patient status and as a unique activity under mobility. Nurses can set additional alarms to alert nurses when high fall-risk patients leave the chair unattended.
Live view now visualizes the doors in the room.
Patient selection field in dashboard and room selection field in patients table now have autocomplete search.
We store each patient visit as a separate "patient session" from now on. That allows the dashboard (and patient live view) to always display the most up-to-date data related to the latest session and also opens up new possibilities for complex reporting in the upcoming patient history page.
Rooms and departments have their own view to list. Users with service role now can view create or update rooms and departments.
With this update, we have brought Ouva wayfinding to mobile devices. Scan the QR code the screen and take the same 3D wayfinding experience on the go - no app needed.
This update increases detection of user presence in front of the screens.
Starting with Dutch, Ouva dashboard now is enabled in multiple languages, including specific rules (gender rules, plural rules) and local date format support.
This is the first, versioned release of our platform beta.
We have completed role-based access functionality for our dashboard. This feature is developed with the least-privileged access methodology in mind. Currently, there are two roles: User and Service. The users without the Service role will not be able to access technical functionality such as Force-Leave, Sensor configuration and others.
This update brings a list of improvements to the map and speech interface and makes space for upcoming QR code/mobile release and voice capture improvements. We have replaced the old assistant interface with an improved one, and made several changes to the wayfinding system.
Verbal directions are now displayed on screen along with the route, in order to make it easier for people who prefer readable directions over maps.
For first-time users, clear directions and suggestions are given for how to use the system.
We have created permanent, color-coded wayshowing routes that display routes to chosen destinations. This allows visitors to quickly identify the way to common destinations without having to use voice or mobile directions.
Ouva will now display announcements set for each device on the dashboard. There is currently no limitation on length, but we recommend maximum 140 characters for readability.
Ouva will now display rotating suggestions after a few seconds for users to know how to get started.
Here are all the related changes:
You can now search by staff name. Here are a few examples:
We have added categorizations to paths as below.
In a planned future update, Ouva will let the person know that to get to an unauthorized destination, they need to check in with staff.
We have updated phrases to be shorter.
You can now narrow your search by floor. Here are a few examples:
You can combine it with building names as well.
You can search by room numbers in any available suite. If the room location is known, it will take you there, otherwise it will take you to the suite. Here are a few examples:
When you ask for a suite, it will now show the range of rooms in that suite next to its name.
You can now search by room or suite number. Here are a few examples:
As visitors may enter one building (e.g. from garage) while they need to get to another, they can now ask for a location combined with the building name as below:
If you ask for a specific building, you will now be taken to its lobby (if you are not in that building). If you are in the right building, Ouva will let you know.
This update resolves a critical bug that caused the system to crash upon asking a question.
If you are going to a doctor’s office, you can search by the name of the staff. If set up for your facility, Ouva will navigate you directly to the office, otherwise it will take you to front desk for further help.
If you know the suite number you are going to, you can ask Ouva for it.
2261 Market St #4436
San Francisco, CA 94114
United States
© 2023 Ouva, LLC
