Small Robot Company

Small Robot Company is reimagining farming to make food production sustainable.

Using robotics and artificial intelligence, SRC have created an entirely new model for ecologically harmonious, efficient and profitable farming. The combination of the robotics in field as well as the user centric software allows the user to view their field results to the per plant level and make accurate application plans on how to best treat their field

  • The overwhelming feedback from farmers across the UK was that traditional farming was not working as much anymore as it was ten years ago.

    Yields are stagnating, machinery costs are rising, and profits are suffering. Everyone is also facing an environmental crisis, with heavy machinery and chemical use playing a key part in contributing to this.

    Small Robot Company is reimagining farming to make food production sustainable. Using robotics and artificial intelligence, they created an entirely new model for ecologically harmonious, efficient and profitable farming.

    This was called "Per Plant farming".

  • Small Robot Company's way of solving this challenge was to invent hardware that could work in tandem with their software.

    The Hardware (Robot) known as 'Tom' would travel up and down the field and take hundreds of thousand of images and detect millions of weed or crops. It could do this with cm's worth of accuracy and give the farmer an extremely granular representation of every data point on their fields within their farm.

    This was then ingested by SRC's custom AI models to output in field maps that could then be visualised on the online platform. This was so that a user could adjust their spraying of that field to be tailored preciously to the data points that had been photographed previously by Tom.

    SRC also developed a per plant, or in this case weed, system that could preciously, and without chemicals, "Zap" a weed in a field using electricity. This would then negate entirely the use for any form of chemical spraying on that field.

    There would always be a trade of, however, with this more ecological form of weeding as there was a balancing act between efficiently and time taken to process a large field. This is why for the pilot SRC decided to cap their product offering to 20ha's of land per user initially.

  • Role + Responsibilities.
    Senior Product Designer

    • Breaking down current interface to a series of MoSCoW requirements.

    • Redesign of the interface in accordance to user feedback.

    • Engaged and responsible for the agile design sprint planning.

    • Testing ideation with internal stakeholders

    • Responsible for the full end to end product delivery


    Project Types.

    New Product Development

    • Ideation workshops

    • Data gathering / breakdown

    • User testing

    Product Redesign

    • High/Low fidelity wireframing

    • Usability revamps

    Product Maintenance

    • Short sprint feature ideation & implementation

Original interface visual Style

  • SRC was lucky to have their home base situated directly across the road from one of their pilot users.

    This fortuitous circumstance meant that the hardware could be tested, in field, almost daily.

    In terms of the software testing the SRC team used Usertesting.com a lot to gauge the usability of the application, the ease of use with navigation and general colour association between data and interface.

    When specific farmer insights were required the in house SRC team would periodically gather focus groups and engage in on site testing via a tablet to scrutinise the users understanding and usage of the platform.

    With all of the above performed and collated the feedback was written up, ticketed and put into the design backlog.

    We were aware that the interface would require a rather drastic overhaul and as the company had also gone through a large rebranding and marketing push it was also deemed "the right time" to change the interface to reflect the new logo and colour scheme which had performed very well under user feedback.

    While investigation pain points throughout the users journey I captured all of this information via quick notes and attached them to the appropriate screens whilst testing with users.

  • Like many revamps it's always good to get an eye on the competition. There are an abundance of farming companies in the UK and there is no shortage of data and information on what they are currently trying to achieve within the UK market.

    I was tasked with researching these competitors and creating one pagers to present to internal stakeholders. The choices were not limited only to those that used in field hardware, as this is quite niche, but included other forms of data gathering including satellite and drone.

    Below are a couple of example slides that were presented as part of a much larger deck and highlighted company offers, technology used, key features and platform accessibility.

  • The SRC platform has multiple moving parts and required a lot of integration between all of the inhouse hardware and software. It was up to the product team, not only to aid the hardware team in their flow of data to the software team but also enable to software team to display this data in a practical and aesthetically pleasing manner.

    Many a MIRO style workshop was taken with the appropriate teams to outline ideas, flows and practicality of the delivery to the deadlines but also, and maybe more importantly, was able to give internal stakeholders a key insight to what features would have to be stripped to allow the product to stay on time for delivery.

    While these features were never completely binned, as they would add great value to product, they were put into the backlog to be picked up at a later date and were tested to see if they would indeed add value to the product.

  • There was a large array of areas that the SRC platform required an overhaul with. While also trying to keep the branding, styles and features consistent, for external and internal use, there were plenty of aspects that required quick and somewhat dirty low fidelity prototypes.

    While I know that some designers opt to do these in hand drawn pen and paper it's just really not my thing. I can work extremely fast with the design software and I usually find it overall quicker to mock something up, share it with the team and iterate based on feedback.

    While not glamourous in the slightest it's probably worth while posting a few Low fidelity screens to show this as part of my process.

    The screens in question were part of a much larger project to create larger efficiency with the internal admin software that would inevitably speed up internal process and allow our Farm advisors to add new members to their teams on their own rather than have to contact our dev team for these members to be added into the database manually.

  • Just like the above with low level it gets to the point where you need to allow people to understand a vision a lot clearer. It’s all very well a designer, or anyone for that matter, having a vision in their head but ideally this needs to be interpreted as a more understandable visual.

    Mocks ups like the one below can be down fairly rapidly. We used a plethora of mockups at this standard with most icons or assets we had the libraries for represented as a pink square. These would later be numbered, categories and higher fidelity mockups use the correct image. But rather than waste time on that for now pink squares tended to suffice.

  • At my time at Small Robot Company there ended up being an extraordinary amount of screens, path and flows to create for different aspects of the platform,

    One of the largest hurdles with map based interfaces, in general, is that while you want to use the map to highlight key data there is also somewhat limited real estate for other aspects of the user interface.

    Ranging from;

    • Administration page

    • Onboarding flows

    • User application maps

    • User profile pages

    • Farm over views

    • Readable field lists

    • Headers with search bar incorporation

    • Full page data analysis comparison

    The SRC platform contained a multitude of different aspects, while consistent, used a variety of layouts depending on the current user flow. This makes map based interfaces a lot more complicated and, hopefully, easy to use as the end result.

Additional material showcasing a large scale prototype for a comparison feature. All created with Adobe XD