LabTrove provides a highly flexible electronic notebook and data management system facilitating the capture of information and the use of this information in a collaborative environment. We have implemented a system to automatically capture, i.e. Blog, information from instruments (Blogjects) and this hugely enhances and improves the collective utilisation and interpretation of the data generated. The system provides for user added semantics (as Key/Value pairs) and the use in the biolabs projects (e.g.http://biolab.isis.rl.ac.uk and http://blogs.chem.soton.ac.uk/neutral_drift) demonstrate how this can be used to great effect in linking processes, data, and analysis (one post per item) and facilities the use of templates (describing the way experiments or other processes should be recorded and facilitating the links between for example reactants, equipment, products, and resulting data). The internal XML structure of the LabTrove system has enabled a link with MIT Simile projects to provide the pivoted TimeLine and Exhibit views that make data recall from the blog and its use as a project management tool very easy.
In addition to use as a lab notebook and data management system LabTrove is also being used to track SANS (Small Angle Neutron Scattering) data reduction in the Mantid framework and provide an audit record of data analysis procedures that can be made available to users. This functionality is currently being used as an exemplar within the WebTracks project to provide a means of automated linking of downstream processed data to the upstream raw data and to provide the beginnings of a route towards linking the experiments and data analysis carried out within the facility to further downstream processing carried out by users in their home institutions. LabTrove in its current form can provide useful functionality at a range of places in our institutional workflow and serves as the glue between a number of currently disconnected processes. Our ultimate aim, which will be supported through the current proposal, will be to use the linking and identification functionality, along with enhanced semantics provided via SPAR (Semantic Publishing and Referencing) and oreChem to enable the creation of full experimental and analytical provenance trails for our experimental processes.
The existing system is deployed or hosted for groups from the USA, through the UK to Australia. The ‘seed corn’ funding from UoS together with OMII resources resulted in the current Open Source product being made available from Sourceforge and via the LabTrove Web site (http://labtrove.org)where the software is available, together with a user manual, and an installation manual (for Debian systems). Alternatively the system is hosted as a VM based system (NGS or Southampton) or provide VM for installation at other data centres. Examples of the use of LabTrove can be seen online as the open science projects can viewed at the http://ourExperiement.org site, some at http://blogs.chem.soton.ac.uk and http://biolab.isis.rl.ac.uk.
The existing LabTrove software has served us well and has provided a good framework in which to understand the interface design and integration issues required for laboratory recording systems and services. However a weakness of the current system, particularly with respect to larger scale integration is the lack of support for existing external vocabularies and data models. A combination of the lightweight approach we have developed along with native support for existing and appropriate structured vocabularies offers an exciting route forwards as we work towards embedding these records in the wider linked data web. Blog3 offers an opportunity to build on our experience and to create a modern tool with the flexibility to serve in a wide range of settings from automated recording to conversation around research objects and integration with a greater range of tools.