DeepScale

Overview

DeepScale AI is a computer vision startup focused on real time automotive perception systems that give cars self driving capabilities.  The team deploys onto cars pre-trained neural networks able to take information from any type of sensor system (including cameras, LiDar, etc.).  They are focused on "efficient" deep learning that is high powered but doesn't rely on expensive and specialized hardware.

DeepScale has a standard software licensing model that charges per sensor and per car to the manufacturer.  Their technology has already been licensed by a number of manufacturers and will be in customer cars in the next few years

The team is around 20 employees and is based in California. 

Why I like Them

What most people don't realize is that existing autonomous vehicles such as Waymo's are massively expensive.  The industry isn't cost-sensitive currently but it will have to become extremely cost conscience as it scales.  The current modus operandi for solving deep learning challenges is to throw more expensive hardware at it.  Automotive manufacturing is a low margin business and would never reach close to profitability with autonomous cars cost structure as they are being developed today.

 I like that DeepScale has thought ahead on this and is focused on creating computer vision systems that work on cars existing electronics (known as ECUs) rather than requiring additional specialized and expensive chips to be installed.  Under the current cost structures of automobile manufacturers, autonomous vehicles will not be able to be sold even close to current car prices, so the team's focus on the business side of the technology is particularly noteworthy.  By using systems already in the latest vehicles, they help automotive manufacturers save on cost by doing more with less.  Think of DeepScale as building efficient automotive AI.  

I also like the competitive nature of the product.  Their systems continue to gather data in the field and are able to be continuously updated with upgraded neural networks pushed to cars remotely that have already been sold.

Disclosure:  I have spoken to members of the team.

Rapid Flow Technologies

Overview

Rapid Flow Technologies is an early stage infrastructure startup working on modernizing traffic systems.  Their flagship product is a smart traffic lighting system called Surtrac.  Current modern day traffic light systems are hard programmed around average traffic flow, but are not adaptive, making them extremely inefficient during most of the day.  Surtrac use AI and cameras that continuously study traffic patterns so they can adjust lighting systems on the fly.  For example, if cameras see a backlog of cars at one point of intersection it would automatically make the green light last longer to clear the backlog.  If no car was in the left turn lane, the system could skip the left turn arrow for that cycle.  The system also communicates its data to nearby traffic lights to help them anticipate when they will receive traffic.

The team was started out of Carnegie Mellon University and is based in Pittsburgh, Pennsylvania.  

Why I like Them

I like that they are an AI company building something that quickly adds tangible value and efficiency visible from day one.  Rapid Flow Technology's systems reduce travel time by 25%, reduce wait time at intersections by 40%, decrease stopping by 30%, and reduce overall vehicle emissions by 20%.  Intelligent Transportation Systems (ITS) and smart infrastructure is a high growth area over the next 2 decades driven by the adoption of autonomous vehicles.  Products like Rapid Flow's are the beginnings of a modern traffic management and vehicular system which will make transportation much more efficient than it is currently.

Disclosure:  All information is from publicly available sources, I have not had any contact with a member of the company or its investors.

Msg.ai

Overview

Msg.ai is an early stage startup that offers a SaaS machine learning product for companies to interact with their customers across all channels including SMS, social media, email, etc.  Basically they offer AI in a box targeting real time online customer service across social media and messaging channels.  When the conversation gets to complex for the AI, the system automatically hands it off to a human agent.  Their product works across a wide range of industries including CPG, eCommerce, online retail, etc.

Founded in October 2014, the company is based in San Francisco.  The founder, Puneet Mehta, is a former Wall Street IT leader and IBM engineer.

Why I like Them

Vertically focused specific use cases like this are where I think the interesting applications are of AI over the next 3-6 years.  Startups that find a way to use AI to automate specific human (aka expensive) processes such as digital customer service are interesting.  It's a relatively easy sell when you can show in a few months the savings dropping straight to a company's bottom line and management can reduce headcount in costs centers.

Even more interesting with Msg.ai is the access to their clients' unique proprietary datasets.  Years of one of a kind records of customer service exchanges in the past are datasets no one else in the industry has access to.  As I bring up time and time again on this blog,  with machine learning its less about the algorithms themselves and more about the proprietary data access they can train the models on that no one else can.

Msg.ai has had strong out of the gate success having already signed Sony and Heinz.  Sony reported in the first 3 months it was able to replace 70 human customer service staff due to Msg.ai technology while making their customer support responses faster.

Disclosure:  All information is from publicly available sources, I have not had any contact with a member of the company or its investors.

Emerging Trends: Automated Food

In the last couple of years a number of startups have appeared in the restaurant industry that seek to automate completely the process of preparing and serving everything from casual meals to beverages.  Several of these startups include:

• Eatsa - A new restaurant concept that is mostly automated and defined by customers having zero interaction with the employees.  Food is ordered online or via in store ipads and when completed appears in electronic cubbies that show the customer's name on them when ready.
• Cafe X - A startup that has built and opened in San Francisco a fully robotic coffee shop.  The actual coffee itself has been reported to be higher quality and half the cost of nearby coffee shops.
• Momentum Machines - A team of roboticists building robotics for restaurants including a high end fully autonomous hamburger machine that can replace 3 workers per machine.  Their mission is explicitly not to build machines to help restaurant workers but to completely replace them.  The startup has plans to launch a burger chain in the next several years.  

As cool as these technologies are, it is critical to keep in mind that they will dramatically replace low skilled human labor.  The US government estimates there are currently half a million people employed at fast food and fast casual eateries.  It is looking like over the next 5-10 years almost all of these jobs will be automated away.  Critics say that consumers prefer human interaction but in my experience a number of people actually prefer not having to interact with another human to purchase a product, especially if an automated system is faster.  In fact, to a number of people (myself included) having no human social interaction when getting food is a feature, not a bug, and something some might even pay an additional amount for.  Larger chains that have kiosks in stores report younger customers will actually wait in line to order from the automated kiosk rather than order from the completely available human cashier next to it.

It is obviously impossible to know without seeing their financials but my guess would be the unit economics for a Cafe X like store are great per order served with faster service, more accurate orders, and less management once the store is setup. Automated produced food has already proven to be faster at dealing with lunch rush hours, sanitation, and food quality consistency.  Simultaneously, some of the labor savings costs can be reinvested into the food itself with higher quality ingredient meals available at lower prices.

Large chains such as Chipotle, McDonalds, etc are certainly watching these startups with keen interest to see consumer reactions and salivating at the chance to cut their rising labor costs.  My prediction is within the next 4-8 years lower end eateries will be 80%+ automated for more efficiency at lower cost.  Human service and interaction will be one of the selling points of mid and high end more formal restaurant dining and even that part of the industry will likely become heavily automated.

SoftWear Automation

Overview

SoftWear Automation is a startup in Atlanta founded by a Georgia Tech computer vision professor that is developing machine vision and robotics technologies for sewn apparel manufacturers.  Sewn apparel is still made by hand in emerging market countries because of the extreme difficulties in sewing cloth for machines.  Unlike with metals or plastics machine have difficulty in knowing where pieces of fabric are when trying to sew them due to their softness.  However, using SoftWear Automation’s machine vision software and systems manufacturing robots can now sew clothing like shirts and jeans more accurately and faster than humans.  The company has only raised $3M in funding to date. Pricing for a system ranges from $50k - $100k.

The company was founded by a computer vision expert at Georgia Tech named Steve Dickerson who currently acts as chairman. The current CEO is a serial entrepreneur named K.P. Reddy who has 20 years of technology entrepreneurship experience having started several IT and communications companies with successful exits.

The sewn apparel manufacturing industry is a $100B+ industry in the US alone.  In 2016 it is still nearly completely un-automated with the work done by humans with sewing machines.  For context, according to the International Federation of Robots in 2014, of the 230,000,000 industrial manufacturing robots sold globally, only 300 were for apparel manufacturing.  The challenge of machines sewing clothing is a technically hard problem hence why there are so little offerings in this space.

Why I like Them

I like them because their technology is head and shoulders  above what is currently on the market and they are already showing strong customer traction.

They started selling in October 2015 with customers making recurring orders especially among fast fashion and athletic apparel manufacturers. By all reports they have strong word of mouth with inbound calls daily from US and Asian apparel manufacturers interested in testing their products.  A strong proponent of their technology is American Apparel founder and serial apparel entrepreneur Dov Charney publicly stating he is using Softwear Automation’s technology in his new clothing venture.

By solving the extremely challenging technical problem of creating sewn apparel without the need for human hands, the product is an easy sell to garment makers. SoftWear Automation’s systems lower manufacturing costs, allow faster production times, shorter inventory cycles, removes liability from human labor, and allows greater customization of products (critical as apparel companies have been forced to have exponentially more SKUs in the last few years due to consumer demand).

Disclosure:  All information is from publicly available sources, I have not had any contact with a member of the company or its investors.