Awarded to Kainos Software Ltd

Start date: Monday 29 November 2021
Value: £7,380,000
Company size: large
Defence Science and Technology Laboratory

AI and Autonomous Systems Experimentation Hub: Agile Delivery partner

10 Incomplete applications

5 SME, 5 large

22 Completed applications

7 SME, 15 large

Important dates

Published
Monday 16 August 2021
Deadline for asking questions
Monday 23 August 2021 at 11:59pm GMT
Closing date for applications
Monday 30 August 2021 at 11:59pm GMT

Overview

Off-payroll (IR35) determination
Contracted out service: the off-payroll rules do not apply
Summary of the work
The Agile AI delivery partner will provide a flexible and scalable AI development capability to supplement the in-government expertise provided by DSTL, providing the MOD with the capacity and skills required to deliver AI and AS concepts sustainably and at scale.
Latest start date
Monday 18 October 2021
Expected contract length
Contract duration to end of FY22/23. Option to extend the duration by 25%, this isn't guaranteed.
Location
South East England
Organisation the work is for
Defence Science and Technology Laboratory
Budget range
Total Limit of Liability is £7.38M (ex VAT), comprising FY21/22: £1.2M, FY22/23: £6.180M. This contract provides Dstl the ability to place up to £7.38m worth of work with the chosen supplier. This amount however is not guaranteed to be spent.

The initial task to be let under this contract is for FY21/22 and has a budget of £1.2m. This price will be the evaluated and hourly/day rates will remain firm for the duration of the contract.

There is the option to increase the LoL of the contract by up to 20% extra if required, this however is not guaranteed.

About the work

Why the work is being done
MOD must increase the pace at which new technologies are developed, explored and adopted for use so that we make best use of the opportunities that technological developments offer, and so that we do so at a pace that enables a perpetual technological edge over adversaries. Artificial Intelligence and Autonomous Systems are two critical and related technologies; consequently Defence has established an AI and Autonomous Systems Experimentation Hub to catalyse the integration and adoption of these critical and ubiquitous capabilities at near-development speed for enduring technological superiority.
Problem to be solved
The Experimentation Hub will conduct rapid AI and AS concept development and experimentation. Dstl requires an Agile delivery partner to provide multi-disciplinary teams at various stages of the agile life-cycle to develop concepts to meet the needs of our users. These teams will have experience and expertise in delivering AI-components of digital and physical systems. We welcome bids from partnered suppliers who can collectively provide the required capabilities, including non-traditional defence suppliers.
Who the users are and what they need to do
As a military user, I need to be able to develop and exploit AI to augment my ability to analyse and gain insight from large volumes of diverse data, so I can improve situational awareness and enhance decision making across a wide variety of military tasks. I need to be able to exploit the benefits of increasingly autonomous systems, and be confident that those systems are robust, safe and trustworthy. As a back-office user, I need to be able to exploit the benefits of AI so I can deliver a wide range of business tasks more efficiently and effectively.
Early market engagement
Any work that’s already been done
Existing team
"The supplier will be working with a mixed team of Dstl specialists, stakeholders from across defence and other contracted partners delivering the Experimentation Hub.

Principal engagement on this task will be with:
Head of AI&AS Experimentation Hub
Project Technical Authority for the Hub
Lead Technical Reviewer
Military Advisor
Lead for AI Enabling Services
Lead for AI Concept Development & Experimentation
Lead for Data Exploitation
Lead for Common Toolbox
Lead for Applied Verification, Validation and Vulnerabilities
Lead for Innovation
"
Current phase
Discovery

Work setup

Address where the work will take place
Dstl Porton Down (SP4 0JQ), Dstl Portsdown West (PO17 6AD), with regular travel to stakeholders around the UK, and remote working opportunities.
Working arrangements
"We would like the supplier to become a partner integrated into the AI and AS Experimentation Hub. This will requires some presence at Dstl sites at Portsdown West and Porton Down, though remote working is encouraged where feasible. Some roles, e.g. management, may require more in office time. Working at Dstl sites will be required for classified elements of the work.

Staff working on projects may have need to travel to conduct research and development activities at the user's locations.

"
Security clearance
Security Check level of clearance is required for majority of team. Some personnel with DV clearance would be advantageous.
The supplier is expected to process and hold their own clearances. Clearances can be managed by 3rd parties. The authority will not hold or sponsor clearances.

Additional information

Additional terms and conditions
DEFCON 703 shall apply throughout.

Skills and experience

Buyers will use the essential and nice-to-have skills and experience to help them evaluate suppliers’ technical competence.

Essential skills and experience
  • Demonstrable experience of managing the development of digital services through discovery, alpha and beta stages.
  • Demonstrable experience of delivering digital services in an agile, user centred way in the public sector, utilising GDS standards.
  • Demonstrable experience in the past 2 years of successfully applying AI techniques to heterogeneous data to deliver benefits to users.
  • Demonstrable experience in the past 2 years of the use of open source data science and machine learning technologies, including Python and R
  • Demonstrable experience in the past 2 years of machine learning operations (MLOps) or similar approaches to deliver live AI-enabled services.
  • Demonstrable experience of providing quality assurance of AI products (verification and validation, test and evaluation).
  • Demonstrable experience of applying good practice and standards to generate reproducible work, effectively recording project artefacts, including: Discovery outputs,Alpha developments, handovers of User Centred Design activity across evolving delivery teams.
  • Demonstrable experience of managing multiple multidisciplinary teams.
  • Demonstrable experience of creating an agile environment within a learning organisation in which groups are nurtured and upskilled to become high performing teams.
Nice-to-have skills and experience
  • Demonstrable experience of delivering AI components of Robotic and Autonomous (physical) systems.
  • Demonstrable experience of delivering AI components of Intelligent Digital (virtual) Systems.
  • Demonstrable experience of military environments and working with uniformed personnel.
  • Demonstrable experience of applying data ethics frameworks.

How suppliers will be evaluated

All suppliers will be asked to provide a written proposal.

How many suppliers to evaluate
5
Proposal criteria
  • How the supplier's approach or solution meets our goal including delivery models, quality assurance and application of relevant standards (0.3)
  • Evidence AI-enabled services that have been developed, the approaches adopted and the issues encountered and resolved. (0.25)
  • Evidence on the diversity of AI-enabled services delivered, supporting Intelligent Digital Systems (virtual) and Robotic and Autonomous Systems (physical) solutions. (0.25)
  • How the supplier has identified risks and dependencies and offered approaches to manage them. (0.1)
  • Team structure and resourcing models, including experience of working alongside clients in co-located, integrated teams. (0.1)
Cultural fit criteria
  • Demonstrate with evidence the ability to create effective supporting governance to nurture agile team working and collaboration and learning across multiple teams. (0.2)
  • Demonstrate with evidence working collaboratively as 'one team' with client and other suppliers, or strategies that will be used. (0.2)
  • Demonstrate with evidence being transparent and collaborative when making decisions. (0.1)
  • Demonstrate with evidence better use of data and user centred design best practises. (0.1)
  • Demonstrate with evidence a no-blame culture with encouragement for people to learn from their mistakes. (0.1)
  • Demonstrate with evidence taking responsibility and embracing accountability for their work, and respond to feedback. (0.1)
  • Demonstrate with evidence of sharing knowledge and experience with other team members. (0.1)
  • Focussed on delivery and making a difference for users (0.1)
Payment approach
Capped time and materials
Additional assessment methods
Presentation
Evaluation weighting

Technical competence

60%

Cultural fit

20%

Price

20%

Questions asked by suppliers

1. Is there an incumbent?
There is no incumbent.
2. What level of access will the successful supplier have to the End User in order to determine form, fit and function of the delivered prototypes/algorithms/functionality.
The successful supplier will be expected to work closely with end users and/or their representatives (e.g. Product Owners) to deliver outputs that meet user needs. Specific access arrangements will be agreed on a task-by-task basis.
3. Is the Authoroty able to share any End User Use Cases for the types of business problem it intends to solve using the developed/prototype technologies?
The concepts to be addressed will be determined at appropriate times throughout the contract, so specific use cases cannot be shared at this point. As per the information already provided, the concepts will involve AI components of Robotic and Autonomous (physical) and Intelligent Digital (virtual) systems.
4. How does the Authority plan to run and operate the contract and how are Statements of Work to be agreed prior to commencement?
Within this overarching contract, Dstl wishes to be able to agree individual tasks with the supplier, in a manner agreeable to both parties. More information will be issued to down-selected bidders during the evaluation phase, and agreed with the successful bidder at an appropriate time in the commercial process.
5. What are the specific technologies/tools used in the Authority’s existing environment with which bidders should be proficient? Or is the build of the AI Dev Test sandpit part of the scope for this procurement?
As per the essential skills and experience, the authority requires demonstrable experience in open source data science and machine learning technologies, including Python and R. The build of an AI Dev Test sandpit is not in scope for this procurement.
6. - What common design standards are current in use by the Authority that the incoming supplier will be expected to adopt and use? Have standards already been agreed for data axchange, governance, quality, test and analysis or any other areas (JSP, STANAG, etc.) that may be applicable to this assignment?
Standards to be used will be agree with the authority during the contract.
7. Are you intending to carry forward the stage 1 scoring for the final evaluation? if yes what scoring % will be carried forward?
The scoring from stage 1 will not be carried over to the next stage, scoring for the proposal and presentation are going to be separate from this first stage.
8. Please describe the data volumes against which the algorithms are to be applied, how frequently do they change, total size (i.e. petabyte), access the supplier can expect to the data, applicable interfaces and state of the data – i.e. structured, unstructured, open, system data, etc.
Data volumes, type etc. will depend on the specific concepts to be addressed, which will be determined through the course of the contract. For each concept, the successful supplier will work with MOD colleagues to assess data requirements, availability and standards applicable to that concept
9. Are the cultural fit criteria to be evaluated at this stage of procurement and – if so – how are they to be assessed?
The cultural criteria will be assessed during the next stage of the process. This stage will require you to respond to our essential skills and nice to haves.
10. Can you please advise the timelines of this procurement (e.g. shortlisted companies notified, anticipated contract award)? Also, can you confirm if shortlisted suppliers will have the opportunity to present?
This first stage lasts 2 weeks ending midnight on Monday the 30th of August. Following evaluation, up to 5 suppliers will be invited to provide a written proposal. This will then be followed by a presentation to complete our evaluation. The following timings are anticipated and may be subject to change:

Publish to Digital Marketplace: 16 – 30 Aug
Initial Questions close: 23 Aug
Closing date for applications: 30th Aug
Notify down-selected bidders: w/c 6 Sep
Written Proposal deadline: 26 Sept
Presentations: 4 Oct
Contract Award: mid/late Oct
11. Which roles (position only without straying into GDPR) from your team will be scoring applications for this opportunity?
Bids will be evaluated by the Dstl project and commercial teams.
12. Could the Authority please confirm the contract length in months?
With an anticipated start period of towards the end of October and lasting until end of FY 22/23, that would work out to be between 17 and 18 months. Any delay past October will take away from the duration of the contract as it stands.
13. Are there any adjacent projects or existing suppliers with which the successful supplier of this assignment will have to interface and what are they?
The AI and Autonomous Systems Experimentation Hub, and hence the chosen supplier for the Agile Delivery partner, will need to interface will multiple projects and organisations. Specific interactions for the Agile Delivery partner will be agreed with the authority over the course of the contract.
14. - Could you reconsider the requirement around sponsorship? We were under the impression that the guidance for DOS5 framework is that suppliers should not be discriminated on the basis of having a pre-existing clearance as long as a supplier is willing to go through the process. The rule limits the supply to innovative non-traditional suppliers in an emerging field like AI where many companies are newer.
Suppliers who currently don't hold suitable clearances can bid on this requirement, however you must be willing to go through the process as clearance is essential to work on this contract. There's an urgent need to process classified information up-to Secret upon contract commencement, we require the majority of supplier personnel to have relevant clearances by contract commencement. As-per previously provided information, we welcome bids from partnered suppliers who can collectively provide the required capabilities, giving suppliers without current clearances the ability to partner with those that do. During the contract, Dstl can sponsor clearances for individuals on a case-by-case-basis.
15. 1. Is there a current or preferred incumbent for this?
2. If there is any then please let us know what are the key pain points?
3. Will you also consider SC eligible candidates and sponsor SC and DV for a candidate?
4. Is this requirement inside or outside of IR35?
1. Answered on a previous question.
3. I believe this has also now been answered on a previous question.
4. Assuming with this question you will require contractors off the payroll to work with you on this contract, it will be down to yourselves to manage any IR35.
16. Will there be further information made available following downselection of the 5 companies, or is this the total of available info?
Yes, more information will be provided to the 5 selected suppliers.
17. Will the potential contract value be a part of any evaluation or contract award criteria? For example, will the potential contract value be compared against a submitting company’s turnover and if so how will this affect the scoring of that company’s submission either at this stage or the proposal stage.
The overall price that gets submitted as part of a suppliers Statement of Work (on the next stage) will be what gets scored. This is not compared against a companies turnover. More information around how the price is scored will be provided on the next stage.
18. Due to annual leave, would it be possible to request a short extension to the submission date?
The initial two week tender period is automatically set within the DOS Framework Digital Marketplace, and can therefore not be extended.
19. can you please confirm whether there is a minimum turnover requirement for a company to have to demonstrate in order to be eligible to bid for this contract as a prime?
There is no minimum company turnover to be eligible to bid for this contract.
20. Please could you clarify the requirement within the nice to have section: Demonstrable experience of delivering AI components of Intelligent Digital (virtual) Systems.
Are you referring to ChatBot like solutions or is this a broader capability you are looking for
Intelligent digital (virtual) systems is our term for any digital system using AI, in contrast to a physical system such as an autonomous vehicle. Examples would include systems to process intelligence, surveillance and reconnaissance data into actionable intelligence, a system to assess courses of action on military operations, or a system to optimise maintenance across a vehicle fleet to improve availability and reduce cost. Essentially a system whose output is better quality information – allowing Defence to make better decisions or to increase its efficiency - rather than a physical output.