Frequently asked questions
When was Harald Blue launched?
Straight to the point: July 2020. 3 things happened end of June:
One - It became clear that the pandemic was going into multiple phases and therefore was far from over,
Two - Contact Tracing using Smart Phone Apps has shown very limited adoption:
- People felt that there was a privacy invasion, because the phone contained a GPS, and personal information.
- People wanted to control what they installed on their own phones.
- Apple’s iPhone have/had problems with running Bluetooth in the background.
- Android phones have/had problems with hacked apps.
- Many countries had had to recall their contact tracing app.
- Adoption rate was limited to 10-20%
- Because the various ways we carry our smart phones (in handbags, in trouser pockets, in shirt pockets, forgot at the desk), the signal strength was uneven and could therefore report lots of false positives on distancing (and vice versa).
- Because each phone supplier using various Bluetooth hardware, configure with various BLE version and various transmission power, the signal strength was uneven and could therefore report lots of false positives on distancing (and vice versa).
What makes you able to get this up and running so quickly?
The founders of Harald Blue are the same founders as XinaBox, which is a modular electronics solution used world over in education and industrial IoT. The founders therefore already have access to both hardware, firmware and software development resources, as well as a supply chain in low-cost PCB manufacturing and component sourcing. The first token is actually based on the same chipset as one of the XinaBox modules uses.
What does it require from our company to use the Harald Blue solution?
Harald Blue offers 3 ways of using the contact tracing solutions:
- Quick and Easy: Your company will receive a spreadsheet pre-populated with the token numbers. You simply fill in the name of the token holder, when you issue the token. Should an employee fall ill, you simply look up the employee’s token number on your sheet and email it to our helpdesk. We will then email you back a spreadsheet with the token numbers to individuals that are possibly infected.
- Self-managed: Your company will receive a login to our backend system, where you will have access to all the tokens allocated your company. You can then insert an employee reference such as an employee number. Should an employee fall ill, you simply look up the employee’s token number or your employee refence number and in return you will receive a list with the token numbers and employee reference detail to individuals that are possibly infected.
- Integrated: Your company can integrate your backend HR system with our system using an API. The API is capable of several different standards and allow you to integrate your backend system with the Harald Blue backend system using either a JSON or XML based REST API.
How do you maintain the privacy of the token bearer?
There are several actions taken to ensure the privacy of the bearer:
- The token contains no GPS and have no knowledge of where the contact happened.
- The Harald Blue backend solution holds no information that outsiders can use to identify the individuals. That means that nobody at Harald Blue have any idea who holds the token. For Harald Blue all token holders are “just a number”.
- While we capture the time when a contact to place, that information is never shared outside our system, and is not shared with the company or its token holders. The time is captured in order to figure when the contact took place compared to an individual being sick.
- The data on the token is not visible to anybody outside Harald Blue. The data is transferred using a secure connection to the backend system, similar to you connecting your internet banking over any network. Data from the token to the backend system is fully encrypted.
- By separating the token bearer’s personal information on the company side, and the contact tracing information on the Harald Blue side, we have created a firewall between the information systems, giving the token bearer maximum privacy.
How do we accurately capture the contact information?
A combination of technology and usage will ensure a minimum of false positives:
- Because all the tokens are made with the same technology, by the same company, we know exactly the transmission power, and how to interpret the signal strength. This is a major issue use smart phone apps with variously hardware implementations of Bluetooth.
- We use Bluetooth Low Energy (BLE) version 5+ which has enhanced signal strength measurements, compared to earlier versions.
- We also capture the movement of the token during the contact, to establish the orientation of the token bearer. The hardware is prepared for that, but we are only activating this part, once we have more data to work with. Until then we follow WHO/CDC standard for contact events, which is within 6 feet/2 meters for minimum 15 minutes (https://www.cdc.gov/coronavirus/2019-ncov/php/contact-tracing/contact-tracing-plan/contact-tracing.html)
- Because we suggest the token bearer wears the token on the chest area, either using a lanyard, or a clip-on solution, or in a shirt pocket, and not carrying in the pocket or handbag, we will get a more uniform and accurate measurement.
Initially we will report the contact information based on WHO/CDC standards (or the company’s own requirements), based on:
- Duration of contact event.
- Distance kept during the event.
- Days since the contact took place and an individual is registered as fallen ill.
- Once enough contact event information has been collected from various participating companies, an AI engine will be used to assist in calculating probabilities of infections based on all collected data including movements sensor data.
How do we/you ensure that the token is been used properly?
A combination of technology and culture has to be used to assure correct wearing of the token:
- By having the token on a lanyard or otherwise visible, the control at entry that also check mask wearing, will be able to verify whether a token is worn.
- Training in the effectiveness and the benefits for both the company keep the company open and only reduce operation with much fewer individuals, as well as the benefits of the bearer knowing that their health is being monitored, should convince all to wear their token.
- A report showing if an allocated token is not moving during a day (using the movement sensor), will give allow management to address the individuals not wearing their tokens.
- Reports from the backend system, will also show battery levels and other technical information about the tokens, excluding any contact events.
- Finally reports to show who is adhering to social distancing and who is not, will help complimenting/rewarding individuals who do go the extra mile.
What about others who either have a token from another company or visitors who don’t have a token?
In some companies there will be a mix of employees from various companies working closely together. An example of that is a manufacturing plant where the just-in-time logistics that brings the components right to the feet of the company workers, will be handled by another company. If both companies are wearing tokens from Harald Blue, the HR department will get a report of possible infected staff even though nobody in the company have been reported ill. That can happen if a contact event happened between two token bearers from different companies.
But if a visitor arrives without a token, then a non-allocated token can be issued to a guest for the duration of the guests visit. Should anybody fall ill at the company that affects the guest, the company can the notify the guest. The other way around also works, if the guest calls in and notify the HR department that she/he has tested positive. The HR department can then using the guests visit time look up the correct token.
By motivating suppliers to the company also to wear a token, the contact tracing becomes more effective. With that in mind if a company is the main economy in the area, it is obvious the most of the staff’s children also attend the same school, so as an extension the staff might want to allow their children to wear a token, or convince the schools, to implement the same token system.
How can you get the battery to last that long?
Several technical solutions have been used to ensure long battery life. Currently we do not have an accurate estimated of the battery life, but we are aiming at 9-12 months.
- The radio uses Bluetooth Low Energy, and as the name dictates the radio uses very little power.
- Furthermore, we use the absolute minimum transmission power, also to avoid false positives. That means our radios are out of reach after single digit meters.
- We also use a movement sensor to see if the tag is in use or not. The sensor shows movement in X, Y and Z direction, and will show 1g (earth’s gravity) in the Y direction when the sensor is worn in a lanyard or on a clip-on, and in the Y and X direction if the clip-on is worn in a shirt pocket. But it will show 1g in the Z-direction if the token is laying down. Furthermore, the movement sensor can also tell if the token is actually moving. Using this information, we simply turn off the token when it is not in use.
- Because the contact event has to last minimum 15 minutes, our tokens wakes up for a few seconds every 5 minutes. They scan for other tokens, record events and goes back to sleep. The 5 minutes are controlled so all tokens wake up exactly at the same time. Because of this technique the tokens are asleep 98.5% of the time and use minimal power in that time.
- The tokens uses WiFi to upload the contact events. This happens every 5-7 hours, when the token is in use. The time frame is random, not to overload the WiFi at one specific time. The upload takes a few seconds and contains a compact form of the contact event information which is transmitted using an SSL (HTTPS) connection to our backend servers located at Amazon Web Services (AWS) in Cape Town, South Africa.
Please feel free to contact us at email@example.com.