Weekly Brief: Torc Robotic Self-Driving Car Travels 1,000 Miles


Highlights of AI news this week:

Partnerships punctuate progress in self-driving cars 

Partnerships help to accelerate work on self-driving cars. This week, NVIDIA joined auto industry suppliers ZF and Hella to their effort to bring self-driving systems to market for OEM clients. Nvidia is joining the two to help incorporate its own in-car AI technology in a self-driving system that meets the New Car Assessment Program (NCAP) certification for passenger vehicles, and to also address safety requirements for commercial and off-road vehicles. Hella builds camera systems, radar systems and other related software systems, ZF is one of the leading tier one suppliers in the car industry ….

Ford has formed a new internal unit to further its efforts in self-driving cars. The new Robotics and AI Research team will operate under Ford’s Research and Advanced Engineering department.  Ford’s VP of Research and Engineering and CTO Dr. Ken Washington said AI and robotics will have a huge potential impact over the next decade. The team will work with Argo AI, the startup that Ford took a majority stake in earlier this year, as well as with other partners.  Washington said in the future we’ll see “at least two” separate fleets of self-driving vehicles on the road operated by Ford: one led by Ford’s own team pursuing advanced research, and another led by Argo AI focused on development and testing of a virtual driver system ….

Another new entrant – one with experience – has joined the field of those offering self-driving car technology to consumer car manufacturers. Torc Robotics has been working on autonomous vehicle tech since 2007, when it finished third in the DARPA Urban Challenge. Now Torc is setting its sights on the consumer car market, with a self-driving car project based on its decade of experience, with more than 1,000 miles logged of autonomous driving in recent tests using two modified Lexus RX vehicles. These have been active on roads since February 2017, driving in “all weather” conditions according to Torc, and equipped with Torc’s in-house localization, mapping, navigation and object detection/tracking systems.

One of Torc’s test vehicles performed a demonstration long-distance drive of 1,000 miles from its Virginia HQ to Ford’s Piquette Avenue Plant in Detroit (shown in photo above), birthplace of the original Model T. Torc sees itself as one of a number of partners working together to help consumer automakers be competitive in the self-driving car market.

Sumo Logic Raises $75M Financing to boost its cloud-based log analysis

A number of deals were announced in the past week to help AI products and services companies move to the next level of maturity.  Sumo Logic, the cloud-based log analysis platform, has raised $75 million, as the seven-year old company may be entering the home stretch before an IPO. While CEO Ramin Sayar was not ready or willing to commit to an IPO timeline, he did admit it was the next logical step for the company, and that the size of the investment gives him the capital to build toward that event. The new round brings the total Sumo Logic has raised to $235 million since its founding in 2010. The round was led by Sapphire Ventures and a virtual Who’s Who of Silicon Valley investors including Accel Partners, DFJ Growth, Greylock Partners, Institutional Venture Partners (IVP), Sequoia Capital and Sutter Hill Ventures…

JASK has emerged from stealth mode to announce it has $12 million in the bank and a machine learning technology that automates network monitoring and management for overtaxed security teams. The idea behind JASK’s service is to provide the services of professionals training in cybersecurity, to companies who cannot hire the security experts on their own. JASK will monitor networks and prioritize threats to chief information security officers and their teams based on the severity of the threat. The company was founded by network security veteran Greg Martin, a co-founder of ThreatStream (now Anomali), which was also developing network monitoring technologies. JASK’s latest round was led by Dell Technologies Capital and TenEleven Ventures, with additional commitments from previous investors Battery Ventures and Vertical Venture Partners…

Netherlands-based Harver has raised $8.1 million in a Series A funding led by Insight Venture Partners to scale the adoption of its AI-powered pre-selection platform, TalentPitch. The platform is designed to replace the resume by integrating with a company’s existing HR processes and systems, and applying predictive analytics to improve the recruitment process. This latest raise takes its total funding to $11.4 million. Harver’s customers, in 13 countries, now include Booking.com, Netflix, Zappos, OpenTable, Casper and Adecco …

Tired of messy medical records, and trying to stay on top of the avalanche of relevant clinical trials, and having too many patients, Dr. Karim Galil decided to create an AI system to match those under his care with the best diagnostic and treatment methods available. He called his new system Mendel.ai after Gregor Mendel, the father of modern genetics science. Mendel.ai has raised $2 million in seed funding from DCM Ventures, Bootstrap Labs and Launch Capital to get the project off the ground.

Mendel.ai is similar in many ways to the U.K.-based BenevolentBio, which is focused on skimming through scientific papers to find the latest in cutting-edge medical research. But rather than using keyword data, Mendal.ai uses an algorithm that understands the unstructured, natural language content within medical documents pulled from clinicaltrials.gov. It compares that data to a patient’s medical record, to return a fully personalized match. The system then evaluates the patient’s eligibility for each suggested treatment within minutes, according to Galil. The startup’s product could be attractive to doctors who find it difficult to keep up on the exhaustive amount of clinical data…

UVeye, an Israeli startup that is building computer vision and machine learning technology to help detect security threats by scanning the underside of passing vehicles, has raised $4.5 million in seed funding. The round was led by Ahaka Capital, with participation from angel network SeedIL.

Initially being applied to roadside security — such as stopping car bombs or drugs smuggling — UVeye’s tech claims to be able to analyse any vehicle from underneath to identify and detect threats that would otherwise be concealed to the human eye, even as it is moving, up to 28 MPH. It does this using “strategically angled and synchronized hi-res cameras” to build a 360 degree digital model. Three seconds after a vehicle passes over UVeye’s ground installed device, the system is able to create a 3D model of the undercarriage and provide high-resolution color visuals to help find security risks …

Silicon Valley-based Drive.ai  has announced that Andrew Ng has joined its board of directors. Ng was the chief scientist at Chinese tech giant Baidu until March, and previously founded and led the Google Brain project, an artificial intelligence effort. Drive.ai also revealed plans to launch a pilot test later this year so customers can ride in its self-driving vehicles. To fund this, it raised $50 million from venture capital firm New Enterprise Associates.

Toyota’s Mobility Assistance Robot Wraps N. American Home Trial

Toyota recently wrapped up its first in-home trial in North America of its Human Support Robot platform. The robot, which goes by HSR for short, is one of the mobility assistance bots that the automaker is developing for use in the realm of improving quality of life for everyday people.

Toyota has a range of projects underway targeting improved mobility for users with limited or impaired capacities, including senior citizens. In addition to the HSR, they’re also developing: a wearable robotic leg brace that restores walking capabilities to individuals with lower body paralysis; technology that helps those with sight problems gain improved awareness of their surroundings, and robots that can help with transferring patients under supervised care from bed, to chairs. Closer to its home business, Toyota is also working on a device that can help with entry and exit of car seats for those with limited mobility.

The HSR trial that Toyota just completed in North America was run with U.S. war vet Romy Camargo, who suffered injuries in Afghanistan during his service that left him paralyzed below the neck. The robot is a wheeled affair with visual sensors and an articulating arm appendage, and can assist Camargo and his family by performing simple tasks around the house like opening and closing doors, as well as fetching water bottles and other objects …

Fingervision isn’t much to look at at first glance. It appears as though someone MacGyvered a GoPro case out of some clear food wrap and bits of plastic, attaching the creation to the end of a $25,000 industrial robot. It’s not far from the truth. The system is cheap by design, making what it can do all the more remarkable. Using a jury-rigged combination of off-the-shelf parts, the Carnegie Mellon-designed setup is able to give robots a rough approximation of a sense of touch.

Post-doctoral robotics fellow Akihiko Yamaguchi has posted a series of videos featuring a industrial roboBaxter performing a wide variety of impressive tasks with the Fingervision system mounted on the end of each arm. The industrial robot (somewhat awkwardly) peels a banana in one, and, in another, it responds to the light grazing of a feather, moving each time it’s grazed by the pink fuzz.

Salesforce Launches New AI Tools for Devlopers at TrailheaDX Conference

New products announced this week aim at helping AI penetrate further into business use. Salesforce launched three AI tools for developers at the TrailheaDX developer conference. These algorithms, which fall under the new Einstein Platform Services, enable third-party developers to add Einstein intelligence to applications built on top of the Salesforce platform.

The new services include sentiment and intent analysis, and some sophisticated image recognition analysis tools, that, when properly trained, can count objects and even recognize attributes like color or size. The three services open up all kinds of possibilities for developers to build sophisticated functionality into apps built on top of Salesforce.

The Einstein Intent tool allows programmers to understand the intent of customer inquiries, which could make it easier to automatically route leads, escalate service cases or personalize a marketing campaign through a custom app. This could be particularly useful for prioritizing customer service inquiries …

Driverless AI is the latest product from H2O.ai aimed at lowering the barrier to making data science work in a corporate context. The tool assists non-technical employees with preparing data, calibrating parameters and determining the optimal algorithms for tackling specific business problems with machine learning.

At the research level, machine learning problems are complex and unpredictable — combining GANs (generative adversarial networks)  and reinforcement learning in a never before seen use case takes finesse. The reality is that a lot of corporate today uses machine learning for relatively predictable problems — evaluating default rates with a support vector machine, for example.

But even these relatively straightforward problems are tough for non-technical employees to wrap their heads around. Companies are increasingly working data science into non-traditional sales and HR processes, attempting to train their way to costly innovation.

All of H2O.ai’s products help to make AI more accessible, but Driverless AI takes things a step further by physically automating many of the tough decisions that need to be made when preparing a model. Driverless AI automates feature engineering, the process by which key variables are selected to build a model.

Baidu Acquires Kitt.ai Chatbot Maker to Continue AI Push

China’s search giant Baidu has made another acquisition to continue its push into artificial intelligence, and specifically to help it carve out a place for itself as a platform for developers who want to create chatbots and other services based on natural language technology.

Baidu has acquired Kitt.ai, a profitable startup based out of Seattle that has developed a framework to build and power chatbots and voice-based applications across multiple platforms and devices. The deal has been confirmed to TechCrunch directly by Baidu. It was also announced on stage at Baidu’s developer event in Beijing, confirmed in a blog post from Kitt.ai, and also made public with a short note from Baidu on Weibo. Financial terms of the deal are not being disclosed, a Baidu spokesperson said.

Kitt.ai has been around since 2014, but it appears that it had only disclosed a seed round of funding of an unspecified size as a startup. Its backers were Amazon’s Alexa Fund and the Seattle-based Founders Co-op.

  • Compiled by John P. Desmond, AI Trends Editor