Introducing BBC Box

Full article:

BBC Box is a Databox-powered platform for experimenting with different models of personal data processing that addresses BBC R&D’s core priorities.

The first Box service will be a recommender that imports user data (with permission) from a range of media services, and processes it within the Databox environment to create a user profile. This can then be exported under user control to shape the suggestions offered by an enhanced media/listings application.

This will be the project’s first end to end demo and will create a space to experiment with new approaches to personal data with defined use cases aimed at demonstrating new forms of audience value. Our roadmap over the next 12 months takes us towards more distinctive and ambitious new forms of value use cases, building on our existing work and partnership with the Databox team and collaborative work with colleagues across BBC Design + Engineering.

Databox part of the new UK DRI program

Databox will play a central role as part of the £20m new Dementia Research Institute (UK DRI) Care Research & Technology Centre at Imperial College London to enable people with dementia to live in own homes for longer.

The goal of the new centre is to develop technologies to enable people to live in their own homes for as long as possible, explains Professor David Sharp, Neurologist at Imperial College London and Head of the new centre: “The vision for this centre is to use patient-centred technology to help people affected by dementia to live better and for longer in their own homes.

The 'Healthy Home' being developed in the new centre

Read more about the centre on

Databox Version: 0.5.1 released 

This version contains a number of bug fixes arm64v8 support and some new features including:

  • Arm64v8 support has been enabled and arm images built. Databox now runs on a Pi3b+ (testing and performance improvements needed)
  • fixed bugs #291#289#285 #284 #277
  • Version manifests and allow addition of manifest repos to the app-store driver
  • Manifests for apps and drivers can now provide a full docker image name and tag
  • Changes to the manifest format to enable core-ui to install dependencies at install time
  • Fixed nodejs builds
  • Containers running as root (all apps and drivers now run as an unprivileged user)
  • the first version of the quickstart guide
  • make databox recover from host and docker demon restarts
  • App and Driver crashes are detected and restarted cleanly
  • removed the need for the CM and SDK to bind mount directories on the host
  • updates to the arbiter to fix export token support
  • The export service has been re-enabled

Databox Version 0.5.0 released!

This version contains a rewritten core-arbiter, core-container-manger and a new set of build tools. There is now no javascript in the code of databox. Most of the core databox components now communicate over the ZestDB protocol, the language-specific libraries have been updated to reflect this change.

The databox user interface has been moved for the core-container-manger into its own component core-ui. This uses a new experimental store based API to access data and API endpoints within the container-manager. This enhances security and enables audit logging by default. It also has the benefit that new user interfaces can be developed in the same way as databox apps.

The platform-app server has also been removed in favour of a databox driver that read manifest from a git hub repository (driver-app-store).

With these changes, the core of databox is much more stable and should be easier or extend and develop on in the future.

Have fun, and as always expect bugs and dragons. Please report issues on the main me-box/databox repository.

The Databox Team


Research Associate/Senior Research Associate at the Cambridge University Computer Laboratory

Research Associate £31,604 – £38,833 or Senior Research Associate £39,992 – £50,618

Fixed-term: The funds for this post are available for 24 months.

A Research Associate post is available in the Systems Research Group at the Cambridge University Computer Laboratory for up to 2 years with the possibility of extension. Appointment to Senior Research Associate will be considered for exceptional candidates. We welcome applications from candidates with experience outside academia.

The Systems Research Group provides a supportive and rigorous environment in which to undertake world-leading research in a wide range of topics in computer systems. The group’s outputs are not limited to publications but often also include spin out companies significant successes include XenSource (acquired by Citrix Systems Inc. for $500M in 2007) and Unikernel Systems (acquired early last year by Docker Inc.).

The group holds a portfolio of projects supporting development, experimentation and data analysis in personal networked systems to support Human-Data Interaction including Databox and DADA .

This post will focus on design, development and systematic evaluation of technology prototypes for traffic management of domestic IoT devices. This will entail using technologies such as eBPF and Linuxkit to gather data to feed a range of machine learning algorithms, as well as consuming the results of those algorithms alongside user inputs to dynamically reconfigure network connectivity as device behaviours evolve.

Successful candidates will hold a Ph.D. in Computer Science, or have equivalent skills and experience through non-academic routes, and must be able to evidence:

  • Ability to communicate clearly in English, in both written and spoken forms.
  • Familiarity with the core Internet protocols (e.g., TCP/IP, UDP, DNS), the BSD Socket APIs, and Linux software development.
  • Experience of or aptitude for systems-level (e.g., OS kernel, device driver, assembly level) software development.
  • Experience of or aptitude for rigorous system measurement and evaluation, including experiment design, data capture, and data analysis. This may have been gained in commercial or industrial settings as well as through production of academic papers.

Other desirable characteristics include:

  • Evidence of an excellent publication record, commensurate with level of experience is also desirable. Candidates from outside academia may be able to evidence this by providing examples of rigorous technical writing published or distributed through channels other than academic conferences and journals.
  • Familiarity with machine learning algorithms, techniques and applications.

For more background information on the Systems Research Group, see where further information about the projects mentioned can also be found.

Informal enquiries should be directed to Richard Mortier (

To apply online for this vacancy, please click on the ‘Apply’ button below. This will route you to the University’s Web Recruitment System, where you will need to register an account (if you have not already) and log in before completing the online application form.

Please ensure you upload your Curriculum Vitae (CV) and a covering letter. If you upload any additional documents which have not been requested, we will not be able to consider these as part of your application.

Please quote reference NR16300 on your application and in any correspondence about this vacancy.

The University values diversity and is committed to equality of opportunity.

The University has a responsibility to ensure that all employees are eligible to live and work in the UK.

Research Associate Job at Imperial College London (Data Analytics/Privacy for personal/home data)

Applications are invited for a Senior Postdoctoral Research Associate position on Data mining, Analytics, Privacy Protocols and Machine Learning aspects of Personal and IoT data for the the EPSRC Databox Project ( ) and the EPSRC Defense Against Dark Artefacts project ( ), ideally with a focus on edge computing, Internet of Things, data mining, data visualization, or designing and implementation of and security and privacy protocols and machine learning algorithms.


Job listing information

  • Reference ENG00357
  • Closing date 1 July 2018

Salary £36,800 – £44,220 plus benefits

This is full time fixed term position for 14 months, with high possibility of further extensions.

Duties and responsibilities

  • Design and building efficient and scalable privacy protocols and machine learning algorithms in conjunction with the Databox system
  • Designing intuitive data discovery and visualisation apps for IoT/Personal data
  • Engage with the other team members in design and conduct of experiments assessing the interaction of users with exemplar apps and data
  • Work on a team including the PI, CoIs, and other PDRAs to design the Databox prototype

Further information and job details available via


Databox Version: 0.4.0

Databox Version 0.4.0 is now available for download:

Changes since last version:

Changes to me-box/databox:

  • Add funding/grant information see me-box/databox/pull/250
  • Always update node modules see me-box/databox/pull/247
  • enable broadcast traffic relay see me-box/databox/pull/243
  • Adds the needed authorization headers to the tests see me-box/databox/pull/241
  • Authentication see me-box/databox/pull/239
  • Update year in license see me-box/databox/pull/238

Changes to core repositories:

Changes to me-box/core-container-manager:

  • Fix app driver startup order see me-box/core-container-manager/pull/26
  • Fixes proxy DNS lookup errors see me-box/core-container-manager/pull/25
  • UI Updates see me-box/core-container-manager/pull/24
  • Adds a timeout and makes sure next() is called see me-box/core-container-manager/pull/23
  • Fix broken link #21 see me-box/core-container-manager/pull/22
  • Authentication see me-box/core-container-manager/pull/20

Changes to me-box/core-arbiter:

  • No changes in this version

Changes to me-box/core-export-service:

  • No changes in this version

Changes to me-box/core-network:

  • Enable driver to see broadcast packets see me-box/core-network/pull/5
  • ignore reverse lookup request see me-box/core-network/pull/4

Changes to me-box/core-store:

  • Update to zest 0.0.6 see me-box/core-store/pull/5
  • Update to v0.0.4 see me-box/core-store/pull/4
  • Move to jptmoore/zest:v0.0.4 see me-box/core-store/pull/3
  • Core-store v2 see me-box/core-store/pull/2

Changes to me-box/lib-node-databox:

  • Fix parsing of observe response with spaces in the payload see me-box/lib-node-databox/pull/36
  • Core store v2 see me-box/lib-node-databox/pull/35
  • Update to core-store v2 api see me-box/lib-node-databox/pull/34

Changes to me-box/lib-go-databox:

  • Fixes parseRawObserveResponse for payloads with spaces see me-box/lib-go-databox/pull/9
  • Update to new care-store API see me-box/lib-go-databox/pull/8

Market engagement in the Databox project

A new Report, commissioned by the Horizon Digital Research Institute, looks at the industry and consumer market engagement opportunities of the Databox project. The report looks at market engagement with the Databox concept, as it emerges from a limited number of conversations with project stakeholders, business consultants and industry experts. It seeks to explore its value proposition for individuals, ‘data-rich’ companies and ‘data-poor’ companies.

See the full report on