The official version of this document can be found via the PDF button.
The below content has been automatically generated from the original PDF and some formatting may have been lost, therefore it should not be relied upon to extract citations or propose amendments.
WQ.370/2020
WRITTEN QUESTION TO THE MINISTER FOR HEALTH AND SOCIAL SERVICES
BY DEPUTY K.F. MOREL OF ST. LAWRENCE
ANSWER TO BE TABLED ON TUESDAY 20th OCTOBER 2020
Question
When compiling the business case and choosing a developer for the Contact Tracing Digital App, will the Minister advise what selection process took place, including tenders or competing bids, to determine the right developer that would deliver the most effective app and the best value for money?
Answer
The support of Jersey companies has been very important to the success of the app and is the subject of a dedicated FAQ question on our website https://covidalert.gov.je/pages/app-management.aspx?#section-5 How Jersey companies were involved in the app.
Digital Jersey, along with other stakeholders, has been actively working on smartphone-based digital contact technologies since February 2020. The goals have always been to:
- acquire a working solution based on Bluetooth technology; and
- ensure integration with our existing contact tracing team; and
- ensure that any app would be compatible with other parts of the common travel area, if not further beyond.
From March until early June, NHSx (England) was working in conjunction with Pivotal Labs, a software development company within the VM Ware group, to develop its own software which was initially envisaged as a UK-wide solution based on a so-called "centralised" approach. It was hoped that this software would function for all the health systems within the UK, including Crown Dependencies such as Jersey. This proved challenging due to the variety of independent rules and localised health policies. To ensure eventual compatibility between the UK and Jersey, we began an independent negotiation with Pivotal to adapt the UK product for Jersey. This reached an advanced stage and a quote was produced with a minimum estimate of £750,000.
On the 10 April, Google and Apple announced an effort to produce a harmonised proximity detection system called the Exposure Notification Service (the GAEN APIs). The intent was to streamline the implementation of Bluetooth-based proximity detection across Android and iOS devices and reduce the technical difficulty of developing a proximity-based contact tracing app. Shortly after the announcement by Google and Apple, NHSx announced the cancellation of its project and pivoted towards the so-called "decentralised" solution based on the GAEN APIs. This created the potential to choose from a slightly larger group of software providers which had demonstrable experience with the GAEN APIs, while maintaining the ability for the Jersey app to achieve interoperability with other apps across the UK.
At this stage, five approaches were under active consideration:
- A local provider
- Pivotal – which had begun to produce its own second attempt based on GAEN
- Zühlke – a company based in Munich, which had recently begun work on a GAEN-based solution for England
- Ubique – a company based in Zurich, which has announced work on the SwissCovid GAEN app
- NearForm – which had produced working software for the Republic of Ireland and Gibraltar.
A virtual meeting of interested parties in Jersey was held on 14 April 2020 to discuss the status of the project in light of the NHSx decision, and to seek input from local software developers. This meeting was well attended. Given the complexity of the software and the need for rapid solution, no local developers expressed an interest in producing this on-island – in fact the opposite occurred, with several companies citing reputational damage as a key risk.
After the NHSx project cancellation, Pivotal Labs appeared to be struggling to redirect its resources after several months of heavy work on the original NHSx "centralised" solution and had focused its attention on the US market.
Zühlke and Ubique were approached on two occasions, but no reply to our expression of interest was ever received.
NearForm was the first developer in Europe to produce a working product based on GAEN and had been working with Google and Apple intensely on the first versions of this technology. NearForm was responsive to our approach and provided an extremely reasonable estimate of £170,000 (for build and ongoing support) and a short development cycle of 8 weeks. The efficiencies came in part because NearForm was able to re-use code already developed for ROI (Republic of Ireland), but also because the Contact Tracing Team in ROI used the same management platform for its case work as the Team in Jersey. By avoiding the need to design a new interface to the existing Contact Tracing platform, the complexity and cost of the work was further reduced.
We were not aware of any other independent (i.e. not heavily supported by an academic organisation) software development companies in Europe which had a working understanding or were actively developing a solution, and in the circumstances we were not prepared to use limited resources in investigating further afield. In short, at that time, no other provider that could have matched NearForm for experience with the technology and the ability to move rapidly. It is also worth noting that this has proved to be a sensible choice and, given that Northern Ireland and Scotland now also use technology developed by NearForm, it has engendered a spirit of mutual collaboration in improving the technology across those countries, as well as significantly simplifying work on interoperability across the common travel area.