Trusted Open Models Institute (TOMI)
04 Jul 2018
Yes
-  

 

 

Supporting Innovation in cognitive computing through the Bridging for Innovators programme.

Yes

​​​​​​

 

Scafell Pike iPad 3.jpg

The Hartree Centre Trusted Open Models Institute (TOMI) is a national certification capability for cognitive computing or artificial intelligence (AI)​ models and applications. TOMI provides a structured approach through which application developers can deploy their AI application within the Hartree Centre's high performance computing infrastructure to assess its operation. Developed in partnership with experts from Brunel University London, the TOMI assessment takes into account fundamental criteria, such as data security and accountability.​​

Between the current debate around AI ethics and codification of new rules for data processing linked to General Data Protection Regulation (GDPR), a gap exists to both standardise and make understandable the use of AI via the development of standard models.​Standard models can express the security constraints of an AI application domain to industry- or project-specific end user constraints. 

Where does the Hartree Centre fit in?

Through the Bridging for Innovators programme, we aim to provide a safe environment for businesses to embrace AI, taking into account ethics and data security. Our powerful high performance computing infrastructure and expertise is perfectly placed to optimise, test and certify AI models for wider industry. We also offer active support across Technology Readiness Levels (TRLs) 4-7, including training courses and workshops.

This ensures that AI applications supported by Bridging for Innovators are of the highest possible quality and therefore capable of making maximum impact on industry, the economy and society. TOMI is also open to partnerships with larger organisations to test applications that they intend to deploy, helping them to optimise products and minimise risk.​

​Open access

The configuration of tests, utputs of results classification and certification of the tested models are logged to a public ledger. This enables the validation of AI Models outside of the TOMI environment. Hosted at STFC at a government facility the TOMI aims to be non-bias with key stakeholders from public bodies. TOMI ensures provenance of data using the pub​lic ledger, along with accountability for the models and security criteria set by domain experts.​

Get involved

Want to use TOMI to test applications that your organisation​ is using or get involved and play an active role in the project? Contact us​​.


Contact: