A patent as­ser­tion be­comes com­mer­cial­ly se­ri­ous when claim lan­guage is con­nect­ed to a real prod­uct, process or sys­tem. En­gi­neer­ing ev­i­dence for patent dis­putes pro­vides that con­nec­tion. It es­tab­lish­es what the ac­cused tech­nol­o­gy ac­tu­al­ly does, which tech­ni­cal fea­tures can be ob­served or ver­i­fied, and whether those fea­tures cor­re­spond to the re­quire­ments of an as­sert­ed patent claim.

For a founder or en­gi­neer­ing leader, the is­sue is rarely ab­stract. A sales op­por­tu­ni­ty may be at risk, a prod­uct launch may need to con­tin­ue, or an in­vestor may re­quire a cred­i­ble view of ex­po­sure. The qual­i­ty of the tech­ni­cal ev­i­dence can shape the avail­able op­tions long be­fore a court makes any finding.

What en­gi­neer­ing ev­i­dence must prove

Patent claims de­fine the le­gal scope of pro­tec­tion, but they use tech­ni­cal lan­guage. A dis­pute there­fore re­quires a dis­ci­plined trans­la­tion be­tween the claim and the prod­uct. The cen­tral ques­tion is not whether two prod­ucts ap­pear sim­i­lar. It is whether the rel­e­vant prod­uct or process in­cludes every re­quired claim fea­ture, in­ter­pret­ed in its prop­er tech­ni­cal context.

A use­ful ev­i­dence pack­age iden­ti­fies the as­sert­ed claim, sep­a­rates it into in­di­vid­ual lim­i­ta­tions, and maps each lim­i­ta­tion to ver­i­fi­able facts. Those facts may arise from prod­uct doc­u­men­ta­tion, source code, cir­cuit di­a­grams, test records, man­u­fac­tur­ing in­for­ma­tion, pho­tographs, stan­dards doc­u­men­ta­tion or in­spec­tion of a pub­licly avail­able product.

This dis­tinc­tion mat­ters. Mar­ket­ing ma­te­r­i­al may sug­gest that a sys­tem has a par­tic­u­lar ca­pa­bil­i­ty, yet it may not show how that ca­pa­bil­i­ty is tech­ni­cal­ly achieved. A data sheet may de­scribe a re­sult, while the patent claim re­quires a spe­cif­ic ar­chi­tec­ture or op­er­at­ing step. Con­verse­ly, a small tech­ni­cal de­tail in a man­u­al or con­fig­u­ra­tion file can be­come de­ci­sive where it demon­strates a claim lim­i­ta­tion that broad prod­uct de­scrip­tions omit.

The same dis­ci­pline ap­plies to de­fence. It is not enough to say that a claim does not ap­ply. The ev­i­dence should iden­ti­fy the miss­ing lim­i­ta­tion, ex­plain the rel­e­vant op­er­a­tion of the prod­uct, and show why the dif­fer­ence is tech­ni­cal­ly meaningful.

En­gi­neer­ing ev­i­dence for patent dis­putes starts with the claim

The most ef­fi­cient tech­ni­cal in­ves­ti­ga­tion be­gins with the as­sert­ed in­de­pen­dent claims. In­de­pen­dent claims usu­al­ly con­tain the fea­tures that must be present for in­fringe­ment. De­pen­dent claims add fur­ther lim­i­ta­tions and may be­come im­por­tant where the broad­er claim po­si­tion is uncertain.

A claim chart is of­ten the work­ing struc­ture. It places each claim lim­i­ta­tion along­side the cor­re­spond­ing ev­i­dence, the source of that ev­i­dence, its date, and an as­sess­ment of con­fi­dence. This cre­ates a trans­par­ent record for man­age­ment, patent at­tor­neys and tech­ni­cal experts.

Claim charts are valu­able be­cause they ex­pose gaps ear­ly. A fea­ture may be present in the de­sign but un­avail­able in pub­lic ma­te­r­i­al. A prod­uct may have changed be­tween soft­ware re­leas­es. A com­po­nent sup­pli­er may hold in­for­ma­tion that can­not sim­ply be as­sumed. Each sit­u­a­tion calls for a dif­fer­ent next step: con­trolled test­ing, doc­u­ment preser­va­tion, fur­ther tech­ni­cal re­view, or a care­ful­ly scoped re­quest for information.

The tech­ni­cal analy­sis must also re­spect claim con­struc­tion. Terms such as “con­fig­ured to”, “cou­pled”, “sub­stan­tial­ly”, or “con­trol unit” can car­ry spe­cif­ic mean­ing in the patent con­text. En­gi­neers should not force or­di­nary prod­uct ter­mi­nol­o­gy onto claim lan­guage, and le­gal teams should not in­ter­pret tech­ni­cal op­er­a­tion with­out en­gi­neer­ing in­put. The strongest po­si­tion is built where both per­spec­tives are test­ed against the ac­tu­al technology.

Ev­i­dence must be traceable

A per­sua­sive con­clu­sion is only as re­li­able as its un­der­ly­ing record. Every item of ev­i­dence should be trace­able to a source, ver­sion and date. For in­ter­nal ma­te­r­i­al, this means pre­serv­ing orig­i­nal files, re­vi­sion his­to­ries, re­lease notes and test data rather than re­ly­ing on recre­at­ed ex­pla­na­tions af­ter a dis­pute has begun.

For ex­ter­nal prod­ucts, the chain of cus­tody mat­ters. If a de­vice is pur­chased, pho­tographed, dis­as­sem­bled or test­ed, record when and where it was ac­quired, the prod­uct ver­sion, se­r­i­al num­ber, soft­ware ver­sion and the test con­di­tions. Re­peat­able meth­ods car­ry more weight than in­for­mal observations.

This is par­tic­u­lar­ly rel­e­vant for con­nect­ed prod­ucts. Func­tion­al­i­ty can change through soft­ware up­dates, serv­er-side pro­cess­ing or con­fig­u­ra­tion set­tings with­out a vis­i­ble change to the hard­ware. Ev­i­dence that ap­plies to one re­lease may not es­tab­lish the op­er­a­tion of an­oth­er. A care­ful analy­sis iden­ti­fies the pre­cise prod­uct and time pe­ri­od at issue.

Choos­ing the right tech­ni­cal proof

There is no sin­gle form of ev­i­dence that fits every patent dis­pute. The right ap­proach de­pends on the tech­nol­o­gy, the stage of the dis­pute, the in­for­ma­tion avail­able and the com­mer­cial de­ci­sion required.

Pub­lic ev­i­dence can be enough for an ear­ly ev­i­dence-of-use as­sess­ment. Prod­uct man­u­als, reg­u­la­to­ry fil­ings, tech­ni­cal pre­sen­ta­tions and pub­lished stan­dards may show a cred­i­ble ba­sis for an as­ser­tion or pro­vide an ini­tial view of risk. It is ef­fi­cient, rel­a­tive­ly dis­creet and of­ten suf­fi­cient to de­cide whether fur­ther work is justified.

Where the claimed fea­ture is not pub­licly vis­i­ble, prod­uct ex­am­i­na­tion may be nec­es­sary. Hard­ware prod­ucts can some­times be in­spect­ed through non-de­struc­tive test­ing, mi­croscopy, mea­sure­ment or care­ful­ly doc­u­ment­ed dis­as­sem­bly. Soft­ware and dig­i­tal sys­tems re­quire a dif­fer­ent ap­proach. Test­ing in­puts and out­puts may es­tab­lish cer­tain be­hav­iours, but ob­served be­hav­iour does not al­ways re­veal the in­ter­nal mech­a­nism re­quired by the claim.

In­ter­nal ev­i­dence is of­ten crit­i­cal for a de­fence. De­sign spec­i­fi­ca­tions, ar­chi­tec­ture doc­u­ments, source code, bills of ma­te­ri­als and val­i­da­tion re­ports can demon­strate how a prod­uct works and, just as im­por­tant­ly, how it does not work. The trade-off is con­fi­den­tial­i­ty. Teams should lim­it ac­cess, pre­serve priv­i­lege where ap­pro­pri­ate through their le­gal ad­vis­ers, and avoid dis­trib­ut­ing sen­si­tive tech­ni­cal files more wide­ly than necessary.

A tech­ni­cal ex­pert re­port may be ap­pro­pri­ate when the facts are com­plex, con­test­ed or like­ly to be scru­ti­nised in for­mal pro­ceed­ings. The re­port should not mere­ly re­peat the claim chart. It should set out the tech­ni­cal back­ground, ex­plain the method­ol­o­gy, dis­tin­guish ob­served facts from as­sump­tions, and ad­dress al­ter­na­tive in­ter­pre­ta­tions fair­ly. Over­state­ment weak­ens cred­i­bil­i­ty. A clear ex­pla­na­tion of un­cer­tain­ty is of­ten more use­ful than a con­fi­dent an­swer that can­not be supported.

Pre­serve facts be­fore they change

Ev­i­dence is of­ten lost through or­di­nary busi­ness ac­tiv­i­ty rather than de­lib­er­ate ac­tion. Source code repos­i­to­ries are up­dat­ed. Pro­to­type units are re­worked. Sup­pli­ers re­place com­po­nents. Em­ploy­ees leave. Prod­uct pages are re­vised with­out an archive.

Once a patent dis­pute is rea­son­ably fore­see­able, es­tab­lish a fo­cused preser­va­tion process. Iden­ti­fy the rel­e­vant prod­uct ver­sions, de­vel­op­ment records, test en­vi­ron­ments, sup­pli­er com­mu­ni­ca­tions and per­son­nel. Pause rou­tine dele­tion for ma­te­r­i­al that may be rel­e­vant, while keep­ing the process pro­por­tion­ate to the mat­ter. Preser­va­tion should be di­rect­ed by the spe­cif­ic claims and prod­ucts in­volved, not treat­ed as an un­con­trolled col­lec­tion exercise.

This work also sup­ports com­mer­cial de­ci­sion-mak­ing. A pre­served de­sign his­to­ry may show that a ques­tioned fea­ture was in­tro­duced only af­ter a cer­tain date, or was nev­er in­clud­ed in the ver­sion sold in a par­tic­u­lar mar­ket. It may re­veal a vi­able de­sign-around al­ready con­sid­ered dur­ing de­vel­op­ment. Such facts can change the ne­go­ti­a­tion po­si­tion materially.

Use en­gi­neer­ing find­ings to make decisions

Tech­ni­cal ev­i­dence should lead to a de­fined de­ci­sion, not be­come a re­port that sits unan­swered in a file. De­pend­ing on the find­ings, a busi­ness may con­tin­ue with con­fi­dence, seek a more de­tailed in­va­lid­i­ty re­view, al­ter a fea­ture, pre­pare a non-in­fringe­ment po­si­tion, ap­proach a li­cence dis­cus­sion, or re­assess launch timing.

A de­sign-around de­serves rig­or­ous analy­sis. Re­mov­ing one vis­i­ble fea­ture may not avoid a claim if the re­vised sys­tem still per­forms the claimed func­tion through an equiv­a­lent tech­ni­cal arrange­ment. Equal­ly, a nar­row­ly tar­get­ed change can be com­mer­cial­ly prefer­able to years of un­cer­tain­ty. The an­swer de­pends on the claim word­ing, the avail­able al­ter­na­tives, prod­uct re­quire­ments and the cost of implementation.

Tech­ni­cal ev­i­dence can also im­prove com­mu­ni­ca­tions with the oth­er side. A con­cise, well-sup­port­ed re­sponse that iden­ti­fies a miss­ing claim fea­ture is more ef­fec­tive than a gen­er­al de­nial. Where a con­cern is jus­ti­fied, an ac­cu­rate en­gi­neer­ing analy­sis helps frame re­al­is­tic set­tle­ment, li­cens­ing or re­design choic­es with­out con­ced­ing more than the ev­i­dence supports.

TIPAG de­liv­ers struc­tured tech­ni­cal analy­sis that con­nects patents to the prod­ucts and de­ci­sions that mat­ter. The work can sup­port in­fringe­ment as­sess­ment, de­fence prepa­ra­tion, ev­i­dence-of-use stud­ies and co­or­di­nat­ed in­put for ex­ter­nal patent attorneys.

The prac­ti­cal aim is straight­for­ward: pre­serve the facts, test the claims against the real tech­nol­o­gy, and act be­fore un­cer­tain­ty be­comes an avoid­able com­mer­cial constraint.