Erik Penser Bank Bolagsdag

A presentation of Irisity by CFO Andreas Höye

At Erik Penser Bank Bolagsdag, 11 March 2021, our CFO Andreas Höye gave a brief introduction of the company, a run-through of the security industry as well as a status update and Irisitys’ plans for the future.

About Irisity
Irisity is a Gothenburg based company with approximately 40 employees. About 50% of the employees are focused on development in the R&D department. We have long experience – more than 10 years and have customers in more than 10 countries around the world. Irisity is listed on Nasdaq First North Growth Market, with a market cap just over 135 million euro.

Why are we here?
The security industry is undergoing a large transformation at the moment – from manpower dependent to a more digital industry. This is were Irisity comes in. We develop advanced deep learning powered algorithms with the help of AI and machine learning. With the help of these algorithms we can help the security industry digitalize. We do this by partnering up with global security companies through a SaaS business model.

The asset security value chain
Security companies’ main objective is of course to protect people and assets from different kinds of threats. The way they do this at the moment is usually with the help of surveillance cameras. However, these cameras are usually senor triggered with very poor analytics. This causes a lot of problems for the Security Operations Center (SOC) since a lot of the information received from the surveillance cameras are of poor quality and a lot of false alarms are triggered. The SOC experience big issues in knowing the difference between false and true alarms and therefore have to act on a lot of false alarms. Guards needs to be dispatched to end-customers more often than is needed. This generates a lot of costs for the securities as well as poor customer service.

Introducing advanced DNN analytics (Deep Neural Network)
This is where Irisity comes in. With the help of our algorithms we can create DNN based analytics, which upgrades surveillance cameras. By doing this we create high precision video alarms that are sent from the surveillance site to the Security Operations Center (SOC). This means that the SOC only receive true alarms, making operations data-driven with high efficiency and creates high customer value. Guards only act when something real is happening at the surveillance site.

Three different customer segments: Camera manufacturers
Camera manufacturers are undergoing a large transformation and has been doing so for the last decade. This transformation goes from analog to IP, from server to cloud and now from cloud to edge. This means that they can embed analytics directly in to the camera. This creates a lot of benefits for the camera manufacturers such as; lower bandwidth consumption, reduced latency, better reliability, privacy compliance, cost savings and Google TPU (Tensor processing unit).

Three different customer segments: Security companies
Security companies are also undergoing a large transformation from manpower dependent to sensors and on towards AI. Artificial Intelligence (AI) provides a lot of benefits to security companies such as more accurate information on important events and cost savings at SOC and guarding – enabling SOC to do more with the same resources or the same with less resources. New revenues are also available for the companies through added functions that they previously could not offer. It is possible to leverage on the existing infrastructure of camera assets.

Three different customer segments: Enterprises
Irisity is focusing on the large companies in the end-customer segment. These enterprises is similarly evolving their surveillance from deterring, to forensic, to real-time alerts. Benefits include proactive guardning, cost effectiveness and increased protection through new functionalities. The enterprises are also able to leverage on CAPEX already taken. The end-customer might already have spent a lot of money on installment of cameras on each site. By using Irisity, which is a camera independent supplier, they can upgrade the system in place to state-of-the-art top modern cameras.

Market
The relationships with security companies and camera manufacturers are normally partnerships whereas enterprises are approached as direct sales. Irisity sells though a SaaS business model to all markets. This means that we get paid per license, per camera, per month. One camera can have several licenses connected to it. It is a huge market within all three segments. The current status for Irisity within these three segments is that we have signed partnership agreement with G4S within security companies – the world largest security company, Axis Communications within the manufacturer segment – the worlds second largest camera manufacturer and several security companies around the world on pilot at the moment. Looking at the Enterprises segment we work within several verticals. Shown as an example in this presentation is the Rail vertical where we currently have won a procurement with SL (Stockholm Public Transport) and several pilots ongoing such as Metro Madrid and Metro de Santiago.

Competition
Irisity is of course not alone doing this around the world. However, we feel that we have a really strong position in the market. Companies such as AgentVI in Israel, Istana in Australia and Senstar in Canada are direct competitors to us, that are also developing some kind of algorithms that can be used within the security industry. There are also indirect competitors. As an example the camera manufacturers Axis, Bosch and Mobotix. They also have some kinds of algorithm analytics in their cameras but of course dependent hardware as they can only be used in their own cameras.

Security as a Service
We are working within a lot of different industries, such as Security (SOC), City Safety, Construction, Rail, Infrastructure, Transportation, Education, E-health, Industrial and people metrics. These are industries where we and our partners have identified the need of our analytics. The product IRIS™ is based on customer focused DNN based algorithms, developed to solve specific challenges or problems.

Advantages with the analytics
IRIS™ has a large detection range where an area that usually would need 22 cameras could be covered by only 4 cameras. This results in high cost savings in both recourses and money but also time savings for operators. As mentioned IRIS™ is camera independent and can upgrade existing cameras with poor quality in to top modern state-of-the-art cameras. Both thermal cameras and optical cameras can be upgraded. 

New algorithms – Fall detection
As mentioned we develop specialized algorithms for selected industries. Last year we came out with both rail defender and violent behavior. At the moment are working on IRIS™ Fall detection that can detect when someone is falling or lying down. This algorithm can as an example be embedded into the Axis TPU camera. The cameras can be placed in a home environment to assist elderly to monitor odd behavior related to sickness or fainting. By having the algorithm embedded into the camera there is no need to have a server in place or being connected to the cloud. Body movements are followed by pose estimation. As soon as a person is feeling ill and have to lay down, IRIS™ notices and can send an alarm directly or after a certain amount of time.

Financial situation of the company
In November last year we closed a financing round of 50 million Swedish kronor where we got new investors. The proceeds from that investment round will be used to boost sales and roll-out on all three segments. We will also increase the speed of developing new specialized algorithms that can be used in the market. Of course, you don’t only need a great product, but also a great team to be able to reach the potential for a company. We believe that we have that at Irisity, both in the management team but of course also in the whole organization. We are a team of individuals who both have successfully built companies before as well as individuals that are specialized within the field.