Community Research and Development Information Service - CORDIS

H2020

SENSEI Report Summary

Project ID: 808943

Periodic Reporting for period 1 - SENSEI (Optimizing Retail Operation With Real-Time Customer Activity Intelligence)

Reporting period: 2018-02-01 to 2018-07-31

Summary of the context and overall objectives of the project

The retail sector is one of the biggest sectors in the EU economy, with almost 1 in 10 people working in over 3.6 million retail companies (https://ec.europa.eu/growth/content/strengthening-competitiveness-european-retail-sector_en). Brick and mortar retail (Physical stores) still account for 89% of total sales volume in retail, but an increasing number of physical stores are closing due to the high operational costs and poor in-store experience when compared to e-commerce. Many retailers are struggling to keep their stores relevant in a constantly evolving market as today’s shoppers demand nothing less than an experience in the physical store that needs be as flexible and convenient as it is in digital channels.

Moreover, while Brick and mortar stores have lost 22% of their market value over 10 years, and digital giants such as Amazon has increased its market value by nearly 2,000% and is moving from e-commerce to invest in brick and mortar stores.
This market movements, mean that we are at the verge of a dramatic shift in the retail industry, one that calls for a fundamental rethinking of labour allocation, technology investments and new approaches to customer experience in stores.

The main goal of SENSEI is to empower the stores of the future to be PHYGITAL, id est to be able to merge the emotions of the physical store shopping experience and the efficiency of the digital e-commerce operators. Through an integrated system of cameras, sensors, and AI algorithms that capture shoppers’ in-store activities, analyses them in real-time Sensei offers a seamless shopping experience and the possibility to automate many of the store operations.

A striking aspect of SENSEI technology is that it enables the collection of business intelligence data about customers who visit physical stores, as retailers already collect about customers who are shopping online. For example, when a consumer shops on a website, the retailer can see which products they click on, which products they buy, what they remove from their shopping cart, how long they dwell on a product image, and many other behaviours. In physical stores, it is much harder to obtain such metrics.
For that, SENSEI technology uses Artificial Intelligence (AI) and Machine Learning (ML) to analyse anonymized images and video, making human-like judgments from them, enabling retailers to know what customers look for in a store, what they pick up and keep, what they put down, where they walk in the store, where they spend time standing and what catches their attention.

SENSEI provides retailers with a whole new concept of operating stores more efficiently that boosts the overall shopping experience for customers while eliminating the friction points in-store. This ambition is tightly aligned with the digital transformation undergoing in retail sector, which has been urging and preparing retail managers to understand the relevance and added-value of changing the paradigm of in-store shopping experience.

Sensei Project conclusions of the action and overall objective is for the SENSEI technology to be integrated into any chained retail store around Europe and globally. Just as stores have cameras anyway, we are aiming for stores to also have the SENSEI technology.
The overall conclusion of the action in Phase 1 is to further develop the project and apply for SMEi Phase 2, to make SENSEI the enabler of the concept of the 100% smart store. This means to a) scale up SENSEI for larger market adoption by adding further API’s with the largest ERP’s, supporting further sensor inputs and producing additional KPI’s and statistical models, b) to massively pilot SENSEI with potential clients, and c) to engage in communication and commercialization activities to guarantee the needed maturity for market readiness to launch SENSEI smart store enabler after the end of the project.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

Since the beginning of the SMEi Phase 1 Sensei has evolved substantially and gained stronger market traction, thus the founders have done a spin-off of the project from METAICG to incorporating it in a new company to further develop it and take it to the market.
SENSEI has already attracted venture capital investors such as large European retailers as German METRO AG and Portuguese SONAE, as well as Techstars accelerator (USA) and two angel investors (from Portugal and North America).

During the project, we deepened our market knowledge and through our continuous interactions with our retail investors and other retailers the product evolved though a more sophisticated market fit, enhancing the solution Sensei offers to retailers.

During the first semester of the year Sensei has won a few prizes and distinctions, namely: in March 2018 Sensei entered IMPACT Growth accelerator (a consortium fund funded by the EU) to take promising european startups to the next level and helping them to raise funds, awarding Sensei a grant of 100K€ (no equity).
In June Sensei was distinguished by the Portuguese Business Angel association with the prize of “Highest Impact startup of the year 2018” and the reputed magazine WIRED distinguished Sensei as one of the Hottest startups in Europe (September 18 edition), Sensei was also subject of an article in Bloomberg Business publication. Recently this month Sensei was invited by EURONEXT group to join their Techshare program that identified the most promising European tech companies with the potential to enter the European stock exchange in the future.

During SMEi Ph1 project Sensei evolved substantially and we have developed the following activities that we describe and achieved the following results that we highlight:

1. Technical and product KPI’s achieved: increased the accuracy of the computer vision algorithms, better Performance of GPU, Increase gross margin of hardware and deployment of a Proof of Concept store lab in our office fully working with all our technology pipeline integrated (about to be concluded).
2. Market Assessment and Pilot Partner Search: we have done a market study to understand better our target market and potencial and already have signed letters of interest to deploy pilots in two European union countries with two of our partners, for which we will be applying to SMEi Ph2 cutoff.
3. Business Model and Risk Assessment: we have done several business model simulations to ensure the best market acceptance, we also developed a pricing strategy taking into consideration the created value for the customer and the ROI of our technology deployment for the retailer, which is quite high. We conducted a risk assessment focusing particularly on the risk related with barriers for market introduction, IP protection and freedom to operate, compliance with privacy and security regulations, technology reliability and growth.
4. IP Assessment: we have done an IP strategy report with the support of European patent attorneys and we have identified patentable inventions within our technology and we are currently on the process of drafting patent applications of it;
5. Business Plan and P&L with Sensei financial projections: we have done a detailed business plan and financial projections for the next five years, with the company plan for market entering and the following plan to scale up the business. We had the valuable support of the coach from the SMEi Ph1 to develop some of these activities.

The SME Instrument and SENSEI next steps:
In line with the current stage of development and business goals, we will proceed to an application to SMEi Phase 2 program to refine and perform the last mile of our product development, moving from TRL6 to TRL8, providing SENSEI with all the necessary tools to invest in the international expansion and global uptake. For this, the company has planned to carry out two pilots in different markets that will be further developed in the SMEi Ph2 pro

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

Sensei has achieved important progress and results beyond the State of the Art, namely technical and product KPI’s that were achieved such as:

• increased the accuracy of the computer vision algorithms in detecting people and extracting their pose within a store environment;
• Better Performance of GPU while running the deep network for people detection and pose extraction using a smaller and faster network model
• Increase gross margin of hardware at least 40%: The cost of setting up the technology is one of the main barriers to entry, so one of our key progression metrics is related to reducing the cost of our equipment and hardware which we did by working with a manufacturing company.
• Proof of Concept of a store lab in our office fully working with all our technology pipeline integrated (about to be concluded in the post project)

Also, in the Intellectual Property and Freedom to operate study we have conducted we concluded that project progress leaded us to patentable invention(s) and, given the preliminary search and analysis that was carried out, the probability of stepping over previously patented subject-matter is low as this is a very new field of application, thus a field evolving very rapid. We are now progressing with patent applications of Sensei.

Overall, SENSEI aims to fulfil the grand vision of democratization its concept for mass adoption. The socio-economic impacts of the adoption of SENSEI solution are far-reaching, fundamentally because:

• It is scalable – not requiring greenfield locations with embedded technology;
• The technology promotes smaller footprint stores - near to the consumer, and optimized logistics, leading to improved resource management;
• It brings more efficiency to the shopping routines of individual customers preserving one of people’s most important assets, time. The accumulated time freed by Sensei will be contributing to improve social and professional lives;
• Anonymous consumer behaviour data can be aggregated at city scale to anticipate consumer behaviour enabling city decision makers, consumers, and companies to optimize shopping activities directly; and also mobility, infrastructure, and city knowledge indirectly, representing energy savings, better quality of life.

Related information

Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top