Two Small Caps For Artificial Intelligence Investors

By Tim Boreham | More Articles by Tim Boreham

SenSen Networks (not yet listed)

A few years ago the University of Technology Sydney hired tech boffin Subhash Challa to inject a little entrepreneurial flair into the thinking of its students.

One would have to say he succeeded, because three years later the youthful professor quit to focus on commercialising SenSen, the ‘internet of things’ play based on a platform developed within that centre of learning.

SenSen, which is raising $6.5m of capital ahead of a back door listing, even has existing clients including NSW’s Roads and Maritime Services (RMS), Crown Casino and Abu Dhabi Airport.

SenSen – which apparently means ‘a thousand times a thousand’ in Japanese — is about software that extracts data from video camera in real time. The resulting data means faster decision making and automated processing for customers.

SenSen may well become a friend of investors, but it’s a foe of speeding motorists and parking transgressors.

In 2011, SenSen won RMS as its first commercial customer, to operate its emerging network of ‘point to point’ cameras (motorists are booked if their average speed over a given interval exceeds the speed limit).

“We came from nowhere to win a significant slice of that market,” Challa says. Listed speed camera operator Redflex is a notable rival in that market. SenSen then modified the product for parking enforcement and it is used by a number of local councils.

The company also monitors double yellow-line offences for the Singapore’s Land Transport Authority and has the mandate for camera-based security at Abu Dhabi Airport, as well as Colonial First State’s local shopping centres.

It also counts passengers getting on and off Sydney’s ferries.

Challa says the company had so many proposals that it decided to focus on the speed and parking markets – before another opportunity knocked.

“Crown heard about us and called us,” Challa says. “They wanted to understand patron and staff behaviour at the tables.”

SenSen modified the product, now called SenGame, to extract data from table cameras such as occupancy rates, hands per hour and bet types. SenGame is used on 200 tables at Crown’s Melbourne money den and is likely to be expanded to Crown’s Burswood facility in Perth.

Given Crown helped developed the product, Crown gets a royalty if other casinos adopt the product.

SenSen’s revenues have been modest to date: around $500,000 in 2014-15, 1.1m in 2015-16 and a likely $2m in 2016-17 (subject to audit).

In the case of the point-to-point stuff, revenues are based around a up-front and ongoing licence fee, while parking is based on an annual licence plus a payment that depends on the volume of infringements. The casino work is based on a monthly per table fee.

The listing is based on the shell of Orpheus Energy, an Indonesian coal explorer that gave up the game citing “less than favourable conditions”.

As with most tech plays, success will depend on who has the most useful mousetrap rather than the most high falutin’ innovation.

The capital raising is based on a $5m general offer and a $1.5m share purchase plan, both at 10c.

The offer closes on September 18 in view of an Oct 3 listing.

Brainchip (BRN) 16.5c

Brainchip is same but different to SenSen, but has a similar interest in gambling (so to speak). In terms of pervasive and all-knowing surveillance, it’s similarly Orwellian in tone.

Brainchip has devised a product, Brain Chip Studio, that will strike fear into anyone with evil intent: camera facial recognition.

Brainchip’s IP is based on the esoteric yet in-vogue notions of spiking neural networks and neuromorphic computing. Ask your teacher about those ones.

An addendum to security cameras, Brain Chip Studio has the ability to recognise thousands of faces in seconds – key appeal to authorities tackling anything from terrorism to soccer hooliganism.

BrainChip is currently being used by the French Homeland Security department – which has a bit of work on its hands these days – and is also being trialled by the French National Police in Toulouse.

In casino settings, the technology has already been used to identify card counters: undesirable punters who always win because they can memorise what cards have been dealt.

The casinos are also interested in detecting the ‘good’ punters and rewarding them for their loyalty.

Brainchip’s technology, devised by entrepreneur and Brainchip chief technology officer Peter Van der Made, is another extension of machine learning that threatens to make clutzy human brains redundant by about next Christmas.

“It’s all about recognising repeated patterns without being taught to do so,’’ says CEO Louis DiNardo. “A face is little more than a pattern.”

Brainchip is involved in discussions or trials with 17 casinos in what’s estimated to be a $250m global market.

But the bigger opportunity lies in anti-terrorism and civil surveillance, a circa $500m market globally.

Brainchip listed in September 2015 in a reverse takeover by Aziana Ltd, raising $4m at 15.7c apiece. It then raised a further $4m in April last year and $6m in a placement last June (at 15c apiece).

So it’s been back to the well more times than a thirsty camel, not that that’s unusual for a minnow.

In the June half Brainchip generated $US369,000 ($482,000) of revenue, mainly from a currency gain and residual oil and gas royalties. The $US133,000 of sales revenue was sourced mainly from the acquisition of French artificial intelligence outfit Spikenet Technology for $790,000 in mid 2016.

With cash of $US4.5m and expected current quarter cash burn of $1.778m, Brainchip looks adequately funded. But it only takes a conventional noggin to work out that Brainchip will likely have to raise more capital in the near future.

About Tim Boreham

Tim Boreham edits The New Criterion. Many readers will remember Boreham as author of the Criterion column in The Australian newspaper, for well over a decade. He also has more than three decades' experience of business reporting across three major publications.

View more articles by Tim Boreham →